Re: "I do consider Ubuntu to be Debian" , Ian Murdock
On Monday 26 March 2007 13:35, Michael M. wrote: > On Mon, 2007-03-26 at 15:28 +1200, Chris Bannister wrote: > > On Thu, Mar 22, 2007 at 05:37:00PM -0700, Michael M. wrote: > > > All that is to say that Ubuntu serves a purpose, and it's a valuable > > > one, IMO. It's not for everybody; nor is Debian, nor any other distro > > > in particular. Ubuntu at least provides an experience quite similar to > > > Debian while doing things that Debian stubbornly refuses to do, like > > > sticking to a schedule. On that score, I agree 100% with Ian Murdoch > > > -- Debian is missing a big opportunity. > > > > What schedule. There was/is no promised schedule. "Dec 6th 2006" was > > never an actual release date. > > The schedule that the release team puts together. It contains target > release dates. Debian missed its December target for Etch. It remains > to be seen whether it will make the new target of 2 April 2007. > > Call it what you want: schedule, timeline, target, whatever. The point > is that the Debian Project doesn't value it enough to stick to it. I > doubt there's a large software project in existance that hasn't missed > its targets sometimes -- Ubuntu, Fedora, openSuSE all have had release > delays in recent memory, and then there's Windows Vista. But Debian is > fairly unique in being so cavalier about it. yes, debian is cavalier about the release date. but every single OS you cited is cavalier about shipping with bugs just to meet the target date (and some other OS not even mentioned. case in point, google for the recent mega-patch for OSX.) i know which i prefer. i have been waiting for etch to go stable for a good year or so. we have some in-house apps that *depend* on a more recent version of the some of the apps/libs. we had to hack some workarounds and it's painful but it works. but i'll take etch when it's good and ready and not a day before. i'd rather have a working OS, free of bugs, late than a half baked, bug-ridden POS, on time. > > Like I said, it's the "when it's ready" attitude taken to the extreme -- > to the exclusion of providing users any kind of predictablility or > expectations of timeliness -- that I don't like. > > > -- > Michael M. ++ Portland, OR ++ USA > "No live organism can continue for long to exist sanely under conditions > of absolute reality; even larks and katydids are supposed, by some, to > dream." --S. Jackson -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: "I do consider Ubuntu to be Debian" , Ian Murdock
On Wednesday 28 March 2007 09:50, Steve Lamb wrote: > Jorge Peixoto de Morais Neto wrote: > > It raises the question. And the answer is yes. > > > > Everybody behaving in a certain way does not make it right. Just as > > everybody thinking something does not make it true. > > Except, of course, when it comes to language, especially idioms, where > a large enough group can make any foul syntax and grammar correct. > Remember, Lexicographers not only define what is "accurate" but what is in > popular use. Hence, "d'oh" now being a word. :P now, isn't that "ironic". -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: apache-mpm-prefork debian perfomance tuning
On Tuesday 27 March 2007 08:44, Roberto C. Sánchez wrote: > On Tue, Mar 27, 2007 at 03:03:11PM +0530, Siju George wrote: > > Hi, > > > > Just wondering for what amount of Load the default configuration of > > apache2-mpm-prefork in Sarge is tuned? > > > > Will it server well about 1,000 users simultaneously? > > It depends. There are so many variables that it is not even funny. so true. > You > probably want to investigate thouroughly what it is that you are trying > to do, figure out what packages you will be using and then look at the > content that you will be serving. Only then will you even begin to have > enough information to make something resembling a guess that is within > an order of magnitude. even then, to answer a question like "how many simultanious users can the site handle" is near impossible. you'd first need to define "simultanious". you may begin to attempt to answer a slightly different question of "how many visitors per hour" or some such. even then, the best guesstimate would be the order of magnitude. simply because there are so many variables involved. i do stress testing every year or so for a cluster of web servers. "httperf", snmp, cactus, rrdb et al. are your friends. enable "response time" logging in apache and keep an eye out for pages taking longer than 0 seconds. (afaik, apache rounds down to the nearest second). enable slow query logging in mysql. then use explain to find out what's causing each 'slow-query' to take as long as it does. fill the database with data before doing your tests. the database behaves differently with a lot of data vs. without (eg. mysql uses table scans even when there are indexes when it figures it has to look at 75% or the data in the table or some such). and if you do end up using a stress generator (for the lack of a better term) like 'httperf', make sure you are not measuring the performance of the machine generating the requests. even the smallest delay you would barely notice when not under stress tends to magnify and contribute to heavy load on the system when it gets busy. so the only way to find out is to really put the system under heavy load using httperf or the like.. hope that helps. > > To put it in perspective, you have something equivalent to "can a > roadway handle 100 vehicles?" Without providing any other information > to help, it is impossible to answer. > > Regards, > > -Roberto -- anoop aryal [EMAIL PROTECTED]
Re: [OT] How to test a bunch of ethernet cards of a cluster?
On Saturday 06 May 2006 13:05, James He wrote: > Hi, all > > My boss wants me to test a bunch of gigabit ethernet cards of a > cluster. He kept getting time-out problems when running some MPI jobs > on the cluster. The problem only happens when the network traffic is > very high (~100MB/s). Therefore, he wants me to determine which > ethernet card(s) is/are having the problem when the traffic is high. > (We don't get any useful information from syslog or the log files of > MPI jobs.) > > I've seen people testing the ethernet card using nc (netcat) -- just > transfer some files using nc and then compare them. Is there any > better way to do this, or any suggestions about some existing > softwares which can automate this, for I have to test a bunch of them? > Thanks a lot. use SNMP to monitor dropped packets, bandwidth utilization etc.. for each one of the machines. (if you haven't used snmp before, you can setup the monitor on one machine to monitor all your machines provided you setup the snmp daemon on each machine you want to monitor) if you have a managed switch, setup snmp daemon on the switch as well and monitor the switch as well. this is not going to help you test it - since i can't tell what may be the problem yet - but might give you enough clues as to a pattern of usage that causes this which you can use to start developing a testing strategy. 'mrtg' is good for tracking bandwidth usage but not much else. i think there is a package called 'cacti' that aims to be a more complete snmp monitoring software. the daemon you need to use used to be called 'net-snmp'. i think it's called 'snmpd' these days. i typically write my own client - depending on what parts i want to monitor and graph - using python and rrd. you might want to pay special attention to the dropped packet related OIDs in either the udp or tcp sections. someone is dropping packets for it to timeout. if you find out who (sender/reciever/switch) you might also find out why. also, tcpdump any ICMP packets. you just might get lucky ;) hope that helps. > > -- > Best regards, > > James He -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: snmp software
On Sunday 14 May 2006 14:46, Robson wrote: > Brad Brock wrote: > >I want to use a software that can manage several PC > >Routers (with debian running on them). What software > >should I use to manage them? Especially to manage > >bandwith, a kind of traffic control among interfaces > >on each router. Thank you. > > > >__ > >Do You Yahoo!? > >Tired of spam? Yahoo! Mail has the best spam protection around > >http://mail.yahoo.com > > You can use MRTG software based on Perl. See http://oss.oetiker.ch/mrtg/ MRTG is only capable of monitoring AFAIK. and is limited to bandwidth utilization. the plus side to that is that is pretty simple to setup. 'cacti' seems to be more full featured. but only for monitoring as well. i haven't found any 'managing' software for SNMP in debian yet. > > Robson > http://bu.bee.pl -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: System Only Recognizes up to 1GB RAM
On Saturday 13 May 2006 11:36, Roberto C. Sanchez wrote: > Martin A. Brooks wrote: > > Mark R. White wrote: > >> Can anyone please help me? > > > > You need a kernel with highmem support included. It may be easiest > > simply to roll your own. > > Actually, it would be easiest to install a -686, -k7, -k8 kernel or > something like that. It is only the -386 kernels (default for install I > believe) which don't have it. the -686 has support for up to 4GB. if you need more, you need to roll your own. > > -Roberto -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Problem with clock
On Wednesday 17 May 2006 15:28, Rodney D. Myers wrote: > On Wed, 17 May 2006 08:15:31 -0600 > > "Joseph Smidt" <[EMAIL PROTECTED]> wrote: > > Ever since I installed unstable the clock never recieves the correct > > time. When I try to resync or restart the computer I get this error > > message: select() /dev/rtc clock tick timed out > > > > Does anybody know what this error is or how I can fix it? > > Joseph Smidt > > > > -- > > - > >Joseph Smidt > > [EMAIL PROTECTED] > > Not necessarily the answer you want, but install ntpdate, and use this; > > sudo /usr/sbin/ntpdate -b -u pool.ntp.org > > I run that as a script once a week, and I keep fairly accurate time. i don't recall what the -b does but you may not want to use ntpdate on a regular basis. ntp-simple or chrony would be a better option. typically, ntpdate should be used at startup (or the very first time you're setting up ntp) and then let ntp slew the time while the OS is running. otherwise you might end up with erratic jumps in time which could screw things up if you have time sensitive apps.. -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: smtp with ssl auth
On Monday 22 May 2006 17:29, Imre Vida wrote: > Hi, > > our smtp server requiers ssl authentication > (user+passwd over port 465) > It refuses to use SASL, and doesn't work with TSL either > although the latter is often refered to as TSL/SSL - > spent quit some time trying to setup postfix properly, and > tried some others (exim, esmtp, msmtp) as well. > > Win applications such as Eudora & co. seem to work well > when username+passwd and "SSL Authentication" is specified. > > Could anybody tell me what exactly is happening on the server > side? What is exactly the protocol that i would need to > setup? out of the box, postfix seems to want to run chrooted (which is great). but saslauthd doesn't know to create the mux file where postfix can find it. edit /etc/default/saslauthd and change the PWDIR line to: PWDIR=/var/spool/postfix/var/run/saslauthd and PARAMS to have a value of "-m $PWDIR" (with the quotes) restart saslauthd and postfix should be able to sasl just fine. make sure a file called 'mux' is at /var/spool/postfix/var/run/saslauthd/ if it still doesn't work, turn on some debugging on postfix and see why it's complaining. > > Thanks > > imre -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: "I do consider Ubuntu to be Debian" , Ian Murdock
On Thursday 29 March 2007 14:55, Steve Lamb wrote: > anoop aryal wrote: > > i'll take etch when it's good and ready and not a day before. i'd rather > > have a working OS, free of bugs, late than a half baked, bug-ridden POS, > > on time. > > Then you'll be waiting forever because even Debian does not ship stable > releases "free" of bugs. touche. but my main argument is that i value debian prioratizing bug-squashing over meeting release dates. and that if debian is guilty of being cavalier about missing release dates, others are guilty of being cavalier about releasing bug-riddled software (knowingly). i switched from windows to redhat (around '96), then from redhat to fedora, then from fedora to debian. now that i'm here, i don't see myself switching to anything else. and that's *because* debian hasn't given in to the temptation of always having the latest and the greatest software and *because* when the new stable is out, i know there'll be minimum surprises. and that's because debian hasn't given in to meeting artificial release dates. as a software dev who does sysadmin only to setup an environment to run the software we develop, i really value the slow/steady releases because otherwise, we'll be on a hamster wheel of upgrades and have to constantly fix our apps to work with the newer system the whole time and not have the time to develop anything new. debian gives us a nice stable target to hit. i do use testing/unstable on my personal machines where i don't mind the occasional breakage - i see that as a chance to send in the occasional bug report to help out. and as a way to prepare for what may be coming down the pipe in the next release. but debian - the way it is - is exactly why i'm using it on servers. so, yeah, given everything else stays the same, i'd take a firmer release date. but not at the expense of getting software that has critical bugs that could be fixed if the release date was moved. after all, if i was really itching for the newer software, all i'd have to do is 'sed -i "s/sarge/etch/g" /etc/apt/sources.list'. not like debian is stopping me from getting the software before it's magical "release" date. just wanted to make sure that the powers that be also hear from people who appriciate the way things are done in debian. -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Tool to monitor system downtimes?
On Thursday 29 March 2007 07:50, Ron Johnson wrote: > On 03/29/07 02:52, Joerg Lange wrote: > > On 3/28/07, Ron Johnson <[EMAIL PROTECTED]> wrote: > >> Parse this single line file: /proc/loadavg. > > > > Thats what I already do in the perl program I wrote, in order to get > > the latest load (first number is the average per 1 minute)... > > Oh, sorry. When you wrote /monitors *only* real downtimes at the > moment/ I though you meant that it did not yet track system load. > > > Is there > > anything else you want to tell me, something that I missed? > > Other than that your mother dresses you funny? Nope. :P > > Seriously, though, I'd insert the records into an SQLite database, > so that you can create all sorts of trend reports. i'd suggest RRD (round robin database) - see the rrdtool package. with any other databases, you have to deal with data accumulation over years of usage. with RRD, it keeps averages where you get detailed info for the present, and less detailed but still useful data for anytime in the past (well, for a reasonable re-definition of 'anytime'). and the file size is always constant. i have all kinds of performance data for the past 3-4 years yet i don't have to worry about archiving old data or overflowing or ending up with garguantian data files. > > > Anyway, it seems that there is not such a tool. cacti and snmp seem to > > be complex, the program from Hugo seems to be for a slightly different > > purpose. All are not command line oriented. > > > > QUESTION FOR ALL OF YOU: > > Would anyone find it benefitial for them what I have in my mind (see > > first email)? If yes, I would continue work and try to put together a > > debian package one day. > > -- > Ron Johnson, Jr. > Jefferson LA USA > > Give a man a fish, and he eats for a day. > Hit him with a fish, and he goes away for good! -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: IBM x346
On Wednesday 11 April 2007 08:51, Ercan Gunes wrote: > Hii all ; > > > > I had a new ibm x346 intel xeon and I want to install debian on ibm > server. But there is a problem with scsi raid . I created raid 1 from scsi > bios. And I start to install new installation. when I came to creating disk > , the raid 1 didn't seem to raid. The kernel see to sparate scsi disk > drive. How can I salve this problem > > > > Did anybody experiences about this before > all the ibm boxes that we have, shipped with a ServRAID management CD (it's built on top of linux live CD). the easiest way to configure them seems to be to use that CD to configure the RAID instead of doing that from the BIOS. we have a couple of x346 boxes and we configured the raid from the CD and never had any problems. btw, you need the 'ips' kernel module. > > > Cherr... -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: iptables, ftp and dnat?
> > It hangs after ls? Sounds like your data traffic gets jammed > > somehow. > > I know I'm jumping in halfway thru the conversation so this might have already been mentioned. But you may want to check if the firewall is blocking ICMP packets preventing PMTU being figured out correctly. The scenerio you're describing sounds too much like the case of 'Fragmentation needed but DF flag is set'. Letting the right ICMP packets (4/3, I believe) thru in your firewall usually solves these problems. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Server for Debian + MySQL
On Mon, 2008-12-08 at 13:18 +, kj wrote: > Adrian Chapela wrote: > > I am thinking on a new server for my mision critical database server. > > This server will have Debian OS and MySQL database server. Requisites: > > 2 CPU (minimun), 32 GB RAM, 2 TB for mysql data files on SAS Hard > > Disks, 500 GB for mysql binlog + system on SAS Hard Disks, RAID with > > two channels (1 for data files and 1 for binlog). > > > > I am thinking in HP , one HP Proliant ML370 but it hasn't 300 GB SAS > > on 2,5". > > > I can't give you advice on specific hardware, but I'd tell you this: > > 1. All other things being equal, get the one with the fastest possible > disc setup. > 2. Make sure you put 64bit Debian on it. > 3. Make double sure you put 64bit Debian on it ;) > Just curious, how big of a difference (indeed, what difference) does 64bit make? We were looking at 64bit for running some of our Java stuff on since the JVM on 32bit can only address so much memory. 64bit was actually slower, at least for Java despite the JVM being able to allocate more memory before being forced to GC. How does something like MySQL behave on a 64bit vs a 32bit platform? > --kj > > - Anoop -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: OT: laptop recomendations
I used to swear by toshiba until 3-4 satallite pros (all bought at around the same time) died. right after the warrenty expired. all of 'em kinda died at around the same time too - within 1-2 months. I hear someone has a howto on how to wave the soldering iron on it to nurse it back to health if you are handy with the said piece of hardware. On one of 'em the hinge broke. After I took it apart to investigate, I was disappointed at the quality of it's chasis. This was the top of the line series too.. No toshiba for me. On Thu, 2008-12-11 at 00:45 +0200, Micha Feigin wrote: > Hello, > > Sorry for being a bit off topic but it's time for a new laptop that will run > linux solely and I'm looking for recomendation on what has a good build > quallity (will travel), descent battery life, although more important is good > computing power and a good screen at 15.4" (needs to be workable with the > screen) at a price range of around 1500$ rough ballpark. Good service is a > must > since it's a working laptop. > > I know that hp and compaq are a big no no (build quality is shaky at best). I > also have the worst experience possible with Sony support on just about every > continent (haven't managed to run into worse). Lenovo 3000 series also has a > bad track run at our uni in terms of build quallity, no experience with the > ideapad pad heard that they are not much brighter. > > Currently the best candidates are the lenovo thinkpad series (either stick > with > the older and probed t61 or go with the t500 or similar), mac (not sure about > the one button issue although the design is nice). > > Runner up is Dell, although the hardware seems a bit cheap when looking at the > drivers (especially the touchpad which tends to be alps which isn't up to par > with the synaptic). > > Toshiba local dealers didn't prove themselves with a friends laptop. > > Can't find anyone with experience with lg and fujitsu. > > Will be happy for feedback/experience/hardware trouble/Service experience in > case of mulfunciton etc. > > Thanks > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: NFS boot with a dhcpless network
On Tue, 2008-12-09 at 23:38 +0200, Micha Feigin wrote: > On Tue, 09 Dec 2008 14:14:23 -0600 > Ron Johnson wrote: [snip] > I can't use a dhcp server on this network. These machines are exposed to the > general network which already has a dhcp server and if I add another one I'm > going to cause trouble over the whole network. > > Besides a dhcp server with the specific setup is going to be a serious pain. > > The machines do have disks which are meant mostly for local user data. The idea > is to have as little as possible in terms of a system on each machine to make > it easier to maintain and keep safe from the users. Looks like the easiest > solution is to have grub and a kernel installed locally on each machine. > > I once was futzing around with 4 really old machines with no PXE support. I was able to get it to boot with a CD with just grub (and just a kernel) on it. Had to use a special stage2 for GRUB in order for it to boot. See here: http://lists.gnu.org/archive/html/bug-grub/2004-04/msg00152.html. Or google for 'GRUB El Torito'. I remember having trouble initially. But once that was figured out, it turned out to be a handy bootloader with a default menu.lst with shell for when menu.lst didn't cut it. Now, I think I put the IP addresses into the menu.lst which then passed it onto the kernel which was built with diskless (NFS root) support. So grub would load the NFSroot kernel on the CD (and I burned almost identical CD - the only difference being a different IP address on the menu.lst file) and the kernel would then proceed with the NFS root filesystem. At least that is how I recall it. It could all be a figment of my imagination/hazy-memory. I might still have the CD lying around somewhere so if you run into trouble email me and I'll look for it. Anoop. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Con[cat]enate two video files
On Tue, 2008-12-23 at 20:01 +, Rodolfo Medina wrote: > Rodolfo Medina wrote: > > >> [...] I tried to `cat' a file with no audio in it with another with audio > >> in > >> it, and what happens is that the audio of the second one is shifted back at > >> the beginning of the `total' file (cat file1 file2 > total). Apart from > >> that, it's fine. But, how to avoid this undesired effect? > > > > "Douglas A. Tutty" writes: > > > Can you add silent audio to the first file first? > > > I tried with Cinelerra, but then I have no audio at all in the final total > file. Maybe mjpegtools can achieve that? Anyone out there with direct > experience with it? > It might have already been mentioned but, I've been using ffmpeg and mencoder (at times, both - since the two seem to be able to do things the other can't). But those are command line tools. PITA when you cannot work off of timelines.. Otherwise, you can splice videos together with offsets etc.. worked quite well for me. .. and do format conversions quite easily. hope that helps. Anoop. > Rodolfo > > -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: Con[cat]enate two video files
On Wed, 2008-12-24 at 03:04 +0100, Vincent Lefevre wrote: > On 2008-12-23 13:28:50 -0600, Anoop Aryal wrote: > > It might have already been mentioned but, I've been using ffmpeg and > > mencoder > > What package? "apt-file search bin/mencoder" finds nothing. > do you have debian multimedia in your /etc/apt/sources.list ? i've got: deb http://www.debian-multimedia.org lenny main I am using Lenny, of course. Anoop. > -- > Vincent Lefèvre - Web: <http://www.vinc17.org/> > 100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/> > Work: CR INRIA - computer arithmetic / Arenaire project (LIP, ENS-Lyon) > > -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Re: forwarding iptable packets
On Wednesday 01 February 2006 01:18 am, Edward Shornock wrote: > On Thu, Feb 02, 2006 at 07:11:07AM +0800, Jon Miller wrote: > > I'm having a major problem trying to forward packets from either a > > workstation of the LAn to the Internet. I want to be sure I'm doing > > this correctly. > > > > I set a forward rule: > > $IPT -A FORWARD -i $INT_IFACE -o eth1 -p tcp --dport 1262 -j ACCEPT > > > > Then I set a PREROUTING rule > > $IPT -A PREROUTING -i $EXT_IFACE -p tcp --dport 1262 -j DNAT > > --to-destination 192.168.xxx.xxx > > > > All I can see using tethereal on the network is SYN packets. > > > > Any idea what I'm doing wrong? > > try > $IPT -t nat -A PREROUTING -i $EXT_IFACE -p tcp --dport 1262 -j DNAT > --to-destination 192.168.xxx.xxx > > The forward rule looks OK. if it's *from* the LAN *to* the internet, shouldn't it be: if you have a static outside IP $IPT -t nat -A POSTROUTING -o $EXT_IFACE -p tcp --dport 1262 -j DNAT --to-source $EXT_IP or, if your outside IP is dhcp assigned $IPT -t nat -A POSTROUTING -o $EXT_IFACE -p tcp --dport 1262 -j MASQUERADE -anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: creating a disk image of a hardware raid
On Wednesday 01 February 2006 11:30 am, listrcv wrote: > Hi, > > I´m looking for a kind of rescue CD I can boot from to create an image > of the contents of partitions residing on a hardware RAID (3ware 7500-4) > --- or to create an image of the whole RAID. The image(s) could be > stored on a disk connected to the IDE port of the board (independent of > the RAID controller) or transfered to a file server on the network. the RAID should be transparent to whatever backs it up. i.e. the hardware (3ware card in this case) only presents a single drive (or whatever the RAID level is) to the OS and consequently the backup software. it's up to you to restore (when and if you need to restore) to whatever (hardware) RAID you want. > > The image(s) should serve as a backup copy from which the system can be > easily restored. It is only needed as a one-time backup to be made > before doing a dist-upgrade of the system. > > Any recommendations are greatly appreciated. > 'mondo' might be what you're looking for. > > GH anoop.
Re: forwarding iptable packets
On Wednesday 01 February 2006 11:32 am, anoop aryal wrote: > On Wednesday 01 February 2006 01:18 am, Edward Shornock wrote: > > On Thu, Feb 02, 2006 at 07:11:07AM +0800, Jon Miller wrote: > > > I'm having a major problem trying to forward packets from either a > > > workstation of the LAn to the Internet. I want to be sure I'm doing > > > this correctly. > > > > > > I set a forward rule: > > > $IPT -A FORWARD -i $INT_IFACE -o eth1 -p tcp --dport 1262 -j ACCEPT > > > > > > Then I set a PREROUTING rule > > > $IPT -A PREROUTING -i $EXT_IFACE -p tcp --dport 1262 -j DNAT > > > --to-destination 192.168.xxx.xxx > > > > > > All I can see using tethereal on the network is SYN packets. > > > > > > Any idea what I'm doing wrong? > > > > try > > $IPT -t nat -A PREROUTING -i $EXT_IFACE -p tcp --dport 1262 -j DNAT > > --to-destination 192.168.xxx.xxx > > > > The forward rule looks OK. > > if it's *from* the LAN *to* the internet, shouldn't it be: > > if you have a static outside IP > $IPT -t nat -A POSTROUTING -o $EXT_IFACE -p tcp --dport 1262 -j DNAT > --to-source $EXT_IP woops, make that: $IPT -t nat -A POSTROUTING -o $EXT_IFACE -p tcp --dport 1262 -j SNAT --to-source $EXT_IP > or, if your outside IP is dhcp assigned > $IPT -t nat -A POSTROUTING -o $EXT_IFACE -p tcp --dport 1262 -j MASQUERADE > > -anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2 applications attached to one serial port.
On Tuesday 31 January 2006 02:32 pm, Rubén Navarro Huedo wrote: > Hello friends: > My name is Ruben and i am from Spain. > I am administrator of an online weather server. > We have a weather station connected to the serial port of the server. > Weather station ONLY sends data, doesn't receive anything from PC. > At this moment we have one software attached directly to /dev/ttyS0, but we > want test another Sotware. > What we have been testing is: > > tee /dev/ptyt0 /dev/ptyt1 < /dev/ttyS0 > /dev/null & > > And the attach one software to /dev/ttyt0 and the other to /dev/ttyt1 > > We have been testing but...some data is losen in /dev/ttyto and /dev/ttyt1 > I don't understand why :( > > Do you know if i could do this using another command? > Am i doing this Ok? > > Please...we need some help, we have been some weeks fighting with it ;( > > A lot of thank's. > > P.D. Excuse my poor English, but i don't explain very well. > > Thank's http://linuxgazette.net/122/TWDT.html#Intro might be of interest anoop.
Re: Force Ethernet NIC to 10baseT
On Thursday 30 March 2006 09:17 am, Erik Dörnbach wrote: > Hello list, > > at a site I have a debian sarge system with 3 NICs running as a firewall, > 2x 3Com905 and 1x 3Com (dunno which exactly now) onboard. All three get > recognized on bootup, share the 3c59x driver (found in the kernel I think) > and can be configured without errors. > > One of the 905s serves as a connection to a duplicate machine in another > building, some 150m down the road connected via some whacky telephone cable > using 2 twisted pairs. This connection has been used with Ethernet and 2 > lame routers just before I put up the 2 debian servers, albeit possibly > with 10baseT. > > Now I don't get a link on the NICs on both ends, let alone any network > communication. > > What I think, I must force both NICs to 10baseT somehow, but I don't > understand quite how to do that when I use 3 NICs in each machine sharing > the same driver. you may want to check out ethtool. 'ethtool -s speed 10' is probably what you're looking for. > > I tried playing with the media option in /etc/network/interfaces, but to no > avail, always giving out errors (not supported parameter and the like, not > in front of that machine atm) upon networking restart... > > Ah yes, don't tell me to get a correct cable, I would do that also... but > it worked before and it's not possible to dig up the whole street! :) > > > Any clues and ideas are really welcome. > > Regards, > > Erik -- anoop [EMAIL PROTECTED]
Re: Using Ubuntu when I'm used to Debian.
On Tuesday 04 April 2006 06:47 pm, Mike McCarty wrote: > Hal Vaughan wrote: > > On Tuesday 04 April 2006 18:15, kamaraju kusumanchi wrote: > >>Hal Vaughan wrote: > >>>Every time someone suggests something > >>>that makes an install easier or an easier config method, there is > >>>always hostility in this group and elsewhere. > >> > >>Do you have some examples? > > > > I've seen and watched it. While it was more true several years ago than > > now, I've watched that kind of ugliness pop up from time to time. > > It permeates the Linux[1] world. > > [snip] > > > past 2 months. It didn't focus directly on user-friendliness, but it > > was a good example of the "I know hi-tech and you don't, so don't > > bother with Linux" attitude. > > This has been a problem with UNIX[2] like operating systems from the > start. It has a cryptic command language, and many arcane facts > must be known in order properly to administer it. It breeds gurus > and guru mentality. > > People who love UNIX almost universally hate operating systems with > a reasonable and well-thought out CLI, like VAX/VMS was. i could never figure out the syntax for path in VMS tho. > The switches in VMS, if applicable, were the same for all > commands. The commands themselves were for the most part > complete words which made sense. > [snip] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Is it possible to simply copy the kernel from one machine to another and use it?
On Friday 07 April 2006 02:39 pm, Yu,Glen [Ontario] wrote: > Hello everyone, > > I was wondering if it's possible to copy the vmlinuz-x.y.z from one machine > to another and have the other machine run properly with it. Here's the > scenario: > > I have 2 systems, both running Debian 3.1 (Sarge), and their hardware is a > little different from each other. Suppose my machine is running a 2.4.x > kernel and the other a 2.6.x kernel, can I simply copy the > /boot/vmlinuz-2.6.x kernel over to my machine's /boot directory, make the > appropriate changes to /boot/grub/menu.lst and have it work as if I had > installed it through apt or dpkg? > > If anyone has actually tried this and got their machines to work, I would > like to know what and how you did it. i am in the process of doing it to a machine where i had to use debootstrap to install the system from an ide drive to a scsi drive (long story). it is pretty easy to get it to work as long as the root filesystem on both machine reside on similar hardware (best-case/simplest-case scenario is both machines use IDE drives to boot). if that's the case, copy the kernel, the modules directory (look in /lib/modules/), the initrd file and you should be set. if the root fs resides on dissimilar hardware (eg. one has IDE and another machine has SCSI or even different SCSI cards etc..) then you'll need to fix your initrd to load the correct modules (or recompile the kernel to inline all needed modules). while this is doable, it is slightly more complicated to give a generic howto. google cramfs and initrd and you should find some discussion on the subject. > > Thanks, > -Glen > > -- > Glen Yu | 416-739-4861 | [EMAIL PROTECTED] > -- -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Is it possible to simply copy the kernel from one machine to another and use it?
On Tuesday 11 April 2006 02:12 pm, Mark Fletcher wrote: > --- anoop aryal <[EMAIL PROTECTED]> wrote: > > On Friday 07 April 2006 02:39 pm, Yu,Glen [Ontario] > > > > wrote: > > > Hello everyone, > > > > > > I was wondering if it's possible to copy the > > > > vmlinuz-x.y.z from one machine > > > > > to another and have the other machine run properly > > > > with it. Here's the > > > > > scenario: > > > > > > I have 2 systems, both running Debian 3.1 (Sarge), > > > > and their hardware is a > > > > > little different from each other. Suppose my > > > > machine is running a 2.4.x > > > > > kernel and the other a 2.6.x kernel, can I simply > > > > copy the > > > > > /boot/vmlinuz-2.6.x kernel over to my machine's > > > > /boot directory, make the > > > > > appropriate changes to /boot/grub/menu.lst and > > > > have it work as if I had > > > > > installed it through apt or dpkg? > > > > > > If anyone has actually tried this and got their > > > > machines to work, I would > > > > > like to know what and how you did it. > > > > i am in the process of doing it to a machine where i > > had to use debootstrap to > > install the system from an ide drive to a scsi drive > > (long story). > > > > it is pretty easy to get it to work as long as the > > root filesystem on both > > machine reside on similar hardware > > (best-case/simplest-case scenario is both > > machines use IDE drives to boot). if that's the > > case, copy the kernel, the > > modules directory (look in /lib/modules/), the > > initrd file and you should be > > set. > > > > if the root fs resides on dissimilar hardware (eg. > > one has IDE and another > > machine has SCSI or even different SCSI cards etc..) > > then you'll need to fix > > your initrd to load the correct modules (or > > recompile the kernel to inline > > all needed modules). while this is doable, it is > > slightly more complicated to > > give a generic howto. google cramfs and initrd and > > you should find some > > discussion on the subject. > > > > > Thanks, > > > -Glen > > -- > > > > Glen Yu | 416-739-4861 | [EMAIL PROTECTED] > > -- > > > -- > > > > > > anoop aryal > > [EMAIL PROTECTED] > > > > > > -- > > To UNSUBSCRIBE, email to > > [EMAIL PROTECTED] > > with a subject of "unsubscribe". Trouble? Contact > > [EMAIL PROTECTED] > > Bear in mind if you do this, your target machine will > have to have hardware that's at least similar enough > to your source machine's hardware that the correct > drivers will be available, either compiled in as > modules or built into the kernel. If the kernel you > wish to copy is a stock kernel, then you are more > likely to get away with this on dissimilar hardware > than if you built your own -- but then if it's a stock > kernel, just dpkg it on the target machine!!! > > Also your target machine will need to have the same or > better CPU to avoid CPU instruction-set assumptions in > the source kernel causing weird effects on the > lower-spec target machine. > > If your hardware is wildly different then this is > going to be troublesome. > > In general it's better to build your kernel on the > machine you intend to use it on, or at least on a > machine that's so close in spec it might as well be > the same machine. Failing that, use a stock kernel > that's been put together by Debian kernel packaging > folks who Know What They Are Doing (TM). These are > designed to run on a wide variety of hardware. good advice. i wish they would also package a kernel for HIGHMEM > 4G. that's been the only reason i've had to compile my own. i don't mind having to compile *all* the modules (as in the default stock kernel). as the whole thing blows by pretty quick on the machines i have to do it on anyway. in fact, i take the stock config, change the memory limit and compile away. but most of the machines i have to do this for are from the same vendor (same arch, CPU etc..) and only differ in components like the RAID controller etc. so i typically build the initrd with all the modules i need and then i can use the same kernel package for all the machines. > > Assuming you're
Re: communication between chrooted mysql, apache, egroupware
On Monday 24 April 2006 00:39, Gezim Hoxha wrote: > Hi everyone. > > I'm wondering what happens when I chroot apache2 (with php module) in > its own jail and then chroot mysql in its own jail too. How will the PHP > scripts be able to connect to mysql? if you use networking in mysql and have your php scripts use TCP connections to connect to mysql, this should be a non-issue. note however, that mysql_connect likes to 'optimize' connections to 'localhost' by using the file 'mysql.sock' as that is much faster - or so the rumor goes ;) if you want to continue using the sock file, you may have to setup some type of symlink or some such so that both jailees can have some way of getting access to that file. much simpler to just enable networking in mysql and use the TCP connections. > > On top of this. I want to chroot egroupware in its own separate jail > too. Is this doable or is it going to get really complicated? again, if you use TCP it becomes a non issue. > > Thanks, > -Gezim -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: New user need some help
On Thursday 27 April 2006 02:47, Magnus Therning wrote: > On Thu, Apr 27, 2006 at 05:57:41AM +, Andrew M.A. Cater wrote: > >On Wed, Apr 26, 2006 at 10:46:52PM -0400, Terry wrote: > >> I just loaded my first Linux machine. It laoded well and I am able to > >> log on with no problem. My question is how do I get the GUI up after I > >> log on? All I get is the prompt. > >> > >> Terry > > > >As root, > > > >apt-get install x-window-system kde kdm > > I simply have to answer this. > > The command line above is clearly not right, it should of course say: > > apt-get install x-window-system gnome gdm um, you're almost got it. really it should have been: apt-get install emacs21 after all, once you're done loading the BIOS that is linux you do want a *real* OS, don't you? > > /M > > P.S. > For the original poster, who might be just a little confused by this, > I'd suggest first trying the live CDs from Knoppix[1] and Ubuntu[2]. If > you like the former then go with Terry's suggestion, if you like the > latter (which would indicate you're an intelligent and sane person) > you'd go with my suggestion. intelligence and sanity are mutually exclusive. :P > > 1. http://knoppix.org/ > 2. http://www.ubuntu.com/ (download the live CD!) PS: i'm only kidding. i so use kde. use gnome too. which reminds me, whatever happened to sawfish? -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] Recipe for a Debian thread that won't die
On Saturday 29 April 2006 11:39, Manaen Schlabach wrote: > I haven't been around this list long but here are the ingredients for > a thread that won't die. Can anyone think of further needed > ingredients? > > Ingredients > > 2 Dozen Broccoli Growers > A heavy dose of green color > 1 Social Contract > 1 smidgen of "how do you address somebody..." > Add politics > > Stir > > Add some light Colour if it still doesn't look right in your ingredients you forgot: 1 spell color with a 'u'. > > Bake -- anoop aryal [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: network managment software
On Thursday 02 February 2006 02:28 pm, Christoph Nenning wrote: > You may have a look at nagios > > regards > > Christoph > > Am Donnerstag, 2. Februar 2006 13:24 schrieb igor: > > Hi ppl, > > > > I need software to make some kind of computer database for my network. I > > need not just ip addresses, but hardware information also, cpu info, hd > > information, capacities, memor modules... > > > > What could you recommend me? 'cacti' might be what you're looking for. > > > > Thanx > > > > -- > > igor <[EMAIL PROTECTED]> > > pletisan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ntp strangeness
On Wednesday 08 February 2006 05:46 pm, Jacob S wrote: > Hello list, > > After I did an apt-get update/upgrade Monday I've had problem keeping > the time set properly on my computer. I have both ntpdate and > ntp-server installed, running a strictly Sid machine. > > Originally I thought maybe it was my motherboard clock, since it was > the first time I had rebooted in a couple months. So I used hwclock to > set both the hardware clock and system clock. Before long the time was > a half hour fast again. I then stopped ntp-server and > deleted /etc/adjtime and /var/lib/ntp/ntp.drift, did you also delete /var/lib/flux.capacitor ? i've been told that that has some weird effect on the space-time-continuum. (sorry, couldn't resist). > thinking it may be > what was throwing the time off. After restarting ntp-server though, it > was only about 30 minutes before my clock was a half hour fast again. > Now it's almost a full hour fast. > > Anybody know what might be causing this? It's about to drive me up the > wall. I never realized how much I look at my computer clock until I > almost walked out the door to an appointment an hour early today. :-) > (At least it's an hour fast and not an hour slow.) > > TIA, > Jacob -- anoop -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdrom scsi problem
On Wednesday 08 February 2006 05:56 pm, [EMAIL PROTECTED] wrote: > I am running Sarge, linux 2.4.27, and grub with a NEC cdrom. I have run > into what seems to be a common problem, but can find no pointer to the > solution. How can I tell the kernel that I want "/dev/hdc=ide-scsi"? Or, > better yet, can I upgrade from 2.4 to a fairly late 2.6 without installing > the distro again? If so, how? > > Sam (replace 'aptitude' with 'apt-get' below if you use 'apt-get' instead of 'aptitude'). aptitude install kernel-image-2.6-686 should do the trick. -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: good dual head x tutorial
On Thursday 16 February 2006 10:20 pm, Gnu-Raiz wrote: > On Thursday 16 February 2006 06:43 pm, > [EMAIL PROTECTED] wrote: > From: > Nelson Castillo <[EMAIL PROTECTED]> > > On 2/16/06, Matt Zagrabelny <[EMAIL PROTECTED]> wrote: > >> hello, > >> > >> i am looking for a good tutorial on how to setup a dual monitor. > >>ive > >> done enough googling and havent seen anything comprehensive. i > >>have > >> found some but they seem to be mostly mandrake(iva) or fedora > >>core or > >>some other distro centric. any suggestions for a distro agnostic > >>tutorial or a debian specific. > > > >I set up two monitors in my home PC and I collected a few > >links. They're not Debian Specific, but they helped me. > > > >http://del.icio.us/arhuaco/multihead > > > >Regards. > > > >-- > >http://arhuaco.org/ > > Those guides are pretty good, but one of the simplest ways to get > dual head monitors is to buy a video card that supports it. Such as > the Matrox, or some Nvidia cards. be very careful of Matrox. i have been using dual head for some years now on a laptop with nvidia card that supports dual heads. works great. then, when i got an SMP box with no AGP or a suitable PCI-Express slot, i thought i'd give matrox cards a shot since i remember them being great for 2-D especially with multiple heads. about $250 later, i can't get dual head to work properly to save my soul. the configuration works, but it locks up randomly. after several days of googling, i read comments from matrox people about how they expect their users to buy commercial drivers that actually work from third parties. and if buying such a driver is out of reach for linux users, tough; since they have *many* price "unsensitive" customers in the DOD etc. so they really didn't care about providing working drivers for the linux market. the problem, it seems, is that the matrox drivers don't like SMP. so it may not be a problem for most people. i, for one, will be staying at a safe distance from matrox. anoop. > > It can be as simple as turning on twinview, and telling xorg, or > xfree what head is what. The Nvidia readme is very good in that > regard, I am using an old gf2 dual monitor card right now. This > card is over 4 years old and still going strong. The original fan > failed, was replaced, and it works great. It's a good feeling to > get some old hardware working for a few years more. > > It also depends on your needs, some people have different needs, > others just want to get into dual head goodness. If you haven't > bought the hardware yet I would consider what exactly you plan to > do. Some want each head as a separate x display, while others want > an extended desktop. It also depends on the window manager you use, > some window managers have problems with xinerama. So it really can > be an art to get things just right. > > I would not worry about a specific distro howto, as what it comes > down to is the detection of the hardware, and your xorg, or xfree > conf file. Most likely you will have to edit it by hand, to add > the proper sections and monitors. Again this all depends on what > level, and job you intend to do. This will also depend on the > hardware you use, as some hardware require their special additions, > and options in the conf file. > > Gnu-Raiz -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Question about exim forwarding
On Monday 27 February 2006 08:57 am, Aaldijk Arno wrote: > Hi, > > My problem is that email forwarded by exim from root to my private email > address still shows "To: [EMAIL PROTECTED]" in its address header, instead > of To: "[EMAIL PROTECTED]" in its To: header. Because of this, some of > my emails do not get delivered. > > How can I configure exim4 to make it work like that? > > I am successfully using gmail as an smtp smarthost, as per the > instructions found here: http://wiki.debian.org/GmailAndExim4 > I want all emails to root forwarded to my email address, and this works > only partially at the moment. > > My /etc/aliases contains the following entry: > root: [EMAIL PROTECTED] > > My /etc/mailname contains: > localhost > dpkg-reconfigure exim4-config and in one of the questions (it might be the first one related to hostname/domainname) you need to put in domain.fi (instead of the default of 'localhost') and it should work. > > > This test mail: echo "my test" | mail -s "test message" > [EMAIL PROTECTED] > > Results in the following message: (domain names changed) > <...> > Received: from localhost ( [62.78.229.194]) > by mx.domain.com with ESMTP id > l22sm36093nfc.2006.02.27.06.46.12; > Mon, 27 Feb 2006 06:46:12 -0800 (PST) > Received: from root by localhost with local (Exim 4.50) > id 1FDjda-Qb-1Z > for [EMAIL PROTECTED]; Mon, 27 Feb 2006 16:46:10 +0200 > To: [EMAIL PROTECTED] > Subject: test message > Message-Id: <[EMAIL PROTECTED]> > <...> > > If I send email like this: > echo "my test" | mail -s "test message" root > > The message does in fact arrive at the destination, but looks like the > following (and triggers spam filters) > <...> > Received: from localhost ( [62.78.229.194]) > by mx.gmail.com with ESMTP id > d2sm3352990nfe.2006.02.27.06.49.30; > Mon, 27 Feb 2006 06:49:31 -0800 (PST) > Received: from root by localhost with local (Exim 4.50) > id 1FDjgn-Ql-4n > for [EMAIL PROTECTED]; Mon, 27 Feb 2006 16:49:29 +0200 > To: [EMAIL PROTECTED] > Subject: test message > Message-Id: <[EMAIL PROTECTED]> > <...> > > Thanks for any help. > > Arno Aaldijk -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: copying a 12GB file
On Wednesday 01 March 2006 09:43 am, Lubos Vrbka wrote: > > FTP would probably be the fastest, but scp (file transfer over SSH) is > > probably already on both machines if they're both Linux. PuTTY is an > > excellent SSH/SCP client for Windows (and Unix) if you need cross > > platform support. > > > > You probably don't want to use SSH Compression as I've seen it actually > > slow down file transfers over fast LAN links. The SSH encryption might > > slow down the transfer as well, so the transfer speed might depend on > > the speed of your CPUs. > > additional comment - iirc, it's possible to choose weak encryption for > ssh that might actually speed it up a bit. to avoid incurring any encryption/decryption overhead, if you are in a trusted network and if both machines have netcat (you probably can get netcat for windows by installing cygwin) you can try the following: on the workstation do: nc -l -p 6000 > my12gigFile and on the mailserver do: nc -q myWorkstationIP 6000 < my12gigFile that is the quickest way to get the file across as far as i know. however, in my experience, the best way to move really large files across is to place the file under a webserver and then use wget with the -c option; since, it seems, that you inevitably run into problems midway thru the long transfer and having the ability to continue where you left off is worth the overhead of http. if you do end up using ftp, look into tftp. hope it helps. > > regards, > > -- > Lubos > [EMAIL PROTECTED]" -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.6 compilation woes...
On Wednesday 01 March 2006 05:32 pm, Doofus wrote: > Hi, > > I hope someone can help me out here. > I'm finding 2.6 quite complex compared with the 2.4 kernels which I > understood and was comfortable with. > > I've recently installed sarge over the net and chose kernel 2.6 to > initially run with. Everything worked well. > I then got the 2.6 source from my debian mirror and did my first > compilation (make menuconfig, make-kpkg clean, make-kpkg > kernel_image, dpkg -i ../*.deb), using the default (and working) .config > file in /boot as a starting point. > > This also went went well even though I still don't fully understand > every aspect of the source tree. > > My problem: > When I choose the newly compiled kernel from the grub boot menu I get, > early in the boot process: > > Kernel panic: VFS: Unable to mount root fs on unknown-block (0,0) > > > I resolved to do some more compiles, only making changes to the default > configuration by one section at a time, in a process-of-elimination > effort to find where the problem is, but even after the first > compilation when the only change I'd made was the processor type, I > still had the same boot failure. > > So, I cleaned the source tree and compiled another kernel using the > completely unmodified .config file that came with the working default > 2.6 sarge installaion. And you guessed it - another kernel panic. > > Now I'm bamboozled. If I compile a kernel using the identical .config > file that was used to compile the working and running kernel and it > won't boot properly, then my powers of fault finding dry up. I'd be > mightily grateful if anyone can give me any ideas as to where the > problem may lie, or other things to try. > > > Configuration: > > Dell Inspiron 8200 P4 laptop > hda1/boot (ext2) > hda2WinXP (ntfs) > hda3FAT32 > hda5/home (reiser) > hda6/ (reiser) > hda7swap > > > Original working and new unworking sections of menu.lst: > > titleDebian GNU/Linux, kernel 2.6.8-2-686 > root (hd0,0) > kernel /vmlinuz-2.6.8-2-686 root=/dev/hda6 ro > initrd /initrd.img-2.6.8-2-686 > savedefault > boot > > > titleDebian GNU/Linux, kernel 2.6.8 > root (hd0,0) > kernel /vmlinuz-2.6.8 root=/dev/hda6 ro > savedefault > boot > > > Thanks again for any help offered. you didn't mention which one was the working stanza and which one was the non working one. you probably need the modules for riser fs either compiled into the kernel (if you don't want an initrd line like in the second stanza), or, you can have it built as a module but make sure mkinitrd knows to create the initrd with the riserfs module in the initrd. you may want to copy /etc/modules into /etc/mkinitrd/modules and tweak it so that you only leave in the modules you need to boot with. -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.6 compilation woes...
On Thursday 02 March 2006 03:41 am, Ramiro Aceves wrote: > On 3/2/06, Doofus <[EMAIL PROTECTED]> wrote: > > anoop aryal wrote: > > >On Wednesday 01 March 2006 05:32 pm, Doofus wrote: > > >>Hi, > > >> > > >>I hope someone can help me out here. > > >>I'm finding 2.6 quite complex compared with the 2.4 kernels which I > > >>understood and was comfortable with. > > >> > > >>I've recently installed sarge over the net and chose kernel 2.6 to > > >>initially run with. Everything worked well. > > >>I then got the 2.6 source from my debian mirror and did my first > > >>compilation (make menuconfig, make-kpkg clean, make-kpkg > > >>kernel_image, dpkg -i ../*.deb), using the default (and working) > > >> .config file in /boot as a starting point. > > >> > > >> This also went went well even though I still don't fully understand > > >>every aspect of the source tree. > > >> > > >>My problem: > > >>When I choose the newly compiled kernel from the grub boot menu I get, > > >>early in the boot process: > > >> > > >>Kernel panic: VFS: Unable to mount root fs on unknown-block (0,0) > > >> > > >> > > >>I resolved to do some more compiles, only making changes to the default > > >>configuration by one section at a time, in a process-of-elimination > > >>effort to find where the problem is, but even after the first > > >>compilation when the only change I'd made was the processor type, I > > >>still had the same boot failure. > > >> > > >>So, I cleaned the source tree and compiled another kernel using the > > >>completely unmodified .config file that came with the working default > > >>2.6 sarge installaion. And you guessed it - another kernel panic. > > >> > > >>Now I'm bamboozled. If I compile a kernel using the identical .config > > >>file that was used to compile the working and running kernel and it > > >>won't boot properly, then my powers of fault finding dry up. I'd be > > >>mightily grateful if anyone can give me any ideas as to where the > > >>problem may lie, or other things to try. > > >> > > >> > > >>Configuration: > > >> > > >>Dell Inspiron 8200 P4 laptop > > >>hda1/boot (ext2) > > >>hda2WinXP (ntfs) > > >>hda3FAT32 > > >>hda5/home (reiser) > > >>hda6/ (reiser) > > >>hda7swap > > >> > > >> > > >>Original working and new unworking sections of menu.lst: > > >> > > >>titleDebian GNU/Linux, kernel 2.6.8-2-686 > > >>root (hd0,0) > > >>kernel /vmlinuz-2.6.8-2-686 root=/dev/hda6 ro > > >>initrd /initrd.img-2.6.8-2-686 > > >>savedefault > > >>boot > > >> > > >> > > >>titleDebian GNU/Linux, kernel 2.6.8 > > >>root (hd0,0) > > >>kernel /vmlinuz-2.6.8 root=/dev/hda6 ro > > >>savedefault > > >>boot > > >> > > >> > > >>Thanks again for any help offered. > > > > > >you didn't mention which one was the working stanza and which one was > > > the non working one. > > > > > >you probably need the modules for riser fs either compiled into the > > > kernel (if you don't want an initrd line like in the second stanza), > > > or, you can have it built as a module but make sure mkinitrd knows to > > > create the initrd with the riserfs module in the initrd. > > > > > >you may want to copy /etc/modules into /etc/mkinitrd/modules and tweak > > > it so that you only leave in the modules you need to boot with. > > > > Thanks for this. > > > > The working stanza was the first one quoted above. The second one was > > added by kernel-package after compiling the new kernel. I'm having > > trouble seeing past the fact that the .config file I'm compiling against > > is *precisely* the one that the working default kernel was built with, > > and yet the new build doesn't work. I thought the kernel-package would > > take care of the boot environment and menu.lst stanzas automatically, > > including any initrd images, but maybe it's not that simple... > > > > I've never had anything to do with ini
Re: lost mysql root password
On Friday 03 March 2006 09:49 am, Matt Price wrote: > Hi folks, > > a while ago I switched over to mysql-5.0, then forgot about it > entirely. Now I'm back to configuring some programs that use mysql, > and... I can't seem to login as root or any other user. > > One option is just that I've forgotten all my passwords (it's been a > while since I used any of the programs directly). Another is that > some configureation issue has reset permissions or something. In any > case I would like to avoid losing all my databases, so I'm wondering > whetherthere's any way to force a reset of theroot password, or > decrypt the permissions table, or whatever. if you installed using apt, you should have a file called /etc/mysql/debian.cnf . you should be able to use that username and password to log in and reset your password. > > I am *not* a cracker, I'm just a little incompetent. > > APpreciate any help you might give. > > Thanks, > > Matt > > > -- > .''`. Matt Price > > : :' : Debian User > > `. `'` & hemi-geek > `- > -- -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: lost mysql root password
On Friday 03 March 2006 02:20 pm, Matt Price wrote: > Thanks everyone for your help, see below for more: > > On 3/3/06, anoop aryal <[EMAIL PROTECTED]> wrote: > > On Friday 03 March 2006 09:49 am, Matt Price wrote: > > > Hi folks, > > > > > > a while ago I switched over to mysql-5.0, then forgot about it > > > entirely. Now I'm back to configuring some programs that use mysql, > > > and... I can't seem to login as root or any other user. > > > > if you installed using apt, you should have a file > > called /etc/mysql/debian.cnf . you should be able to use that username > > and password to log in and reset your password. > > this seemed the simplest of the options presented to me, and > fortunately, voila! it owrked. I logged in, and had to use GRSNT > statements to reset permissions: > > mysql> grant all privileges on *.* to 'root'@'localhost' identified by > 'apassword' with grant option; > > that worked fine. THen I was looking around and I notice that the > user password seemed to have some dublicate rows in it: > +--+--- >---+---+ > > | host | User > | >| Password | > > +--+--- >---+---+ > > | localhost| root > | >| hast 1 > | > | anarres | root > | >| hash 1 > | > | localhost| root > | >| hash 2 > | > | %| root > | >| hash 2 > > +--+--- >---+---+ > > [passwords are of course hidden in the above] > hmm, that's odd I thought. what to do? It got odder when I tried: > > select * from user where User='root'; > > then only the second set of records showed up. on the other hand, > > mysql> select User from user where User LIKE 'root%'; do: select hex(User) from user where User LIKE 'root%'; that should give you the hex values of the characters that are there. > > have all four. I guess there must be some white space in the username > somewhere. Is there an easy way to identify the precise value of a > mysql field (e.g. by dumping to a CSV file)? I'd like to try to figure > out what went wrong, and deletethe defective lines. > > Thanks again for your help, it's so great to have this working. > > Matt > > > > I am *not* a cracker, I'm just a little incompetent. > > > > > > APpreciate any help you might give. > > > > > > Thanks, > > > > > > Matt > > > > > > > > > -- > > > .''`. Matt Price > > > > > > : :' : Debian User > > > > > > `. `'` & hemi-geek > > > `- > > > -- > > > > -- > > > > > > anoop > > [EMAIL PROTECTED] > > > > > > -- > > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > > with a subject of "unsubscribe". Trouble? Contact > > [EMAIL PROTECTED] -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: lost mysql root password
On Friday 03 March 2006 03:12 pm, Matt Price wrote: > > do: > > > > select hex(User) from user where User LIKE 'root%'; > > > > > > that should give you the hex values of the characters that are there. > > > > > have all four. I guess there must be some white space in the username > > > somewhere. Is there an easy way to identify the precise value of a > > > mysql field (e.g. by dumping to a CSV file)? I'd like to try to figure > > +--+ > > | hex(User)| > > +--+ > > | 726F6F74 | > | 726F6F74202020202020202020202020 | > | 726F6F74 | > | 726F6F74202020202020202020202020 | > > +--+ > 4 rows in set (0.00 sec) > > thanks anoop! I guess those 02's are spaces then... Looks like most > of the user lines from my old db are corrupted in this way as well. > wierd. i would check the encoding being used vs. the encoding that was used when it was initially created. it sounds like you were using a wide_char encoding (eg. UTF-8) before and somehow has now reverted back to latin-1 or some other single char encodings. i'm not an expert on encodings etc.. know just enough to be dangerous. but if this is database wide, (look at char/varchar/text fields and they should all display this behavior), this is encoding related. on a wide char encoding (say utf-8), the database reserves multiple bytes per char not knowing what char it will need to save there. when you tell mysql that it's not wide char, it will just show you what it has - including the previously reserved bytes. it's odd that it's using x20 to pad data tho. or something like that. > Thanks much for your help! > > matt -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: lost mysql root password
On Friday 03 March 2006 04:38 pm, Gene Heskett wrote:
> On Friday 03 March 2006 17:08, anoop aryal wrote:
> >On Friday 03 March 2006 03:12 pm, Matt Price wrote:
> >> > do:
> >> >
> >> > select hex(User) from user where User LIKE 'root%';
> >> >
> >> >
> >> > that should give you the hex values of the characters that are
> >> > there.
> >> >
> >> > > have all four. I guess there must be some white space in the
> >> > > username somewhere. Is there an easy way to identify the
> >> > > precise value of a mysql field (e.g. by dumping to a CSV file)?
> >> > > I'd like to try to figure
> >>
> >> +--+
> >>
> >> | hex(User)|
> >>
> >> +--+
> >>
> >> | 726F6F74 |
> >> | 726F6F74202020202020202020202020 |
> >> | 726F6F74 |
> >> | 726F6F74202020202020202020202020 |
> >>
> >> +--+
> >> 4 rows in set (0.00 sec)
>
> Is that even encoded at all? That looks a bit like it would say
> "root", twice, in ascii to me. Look it up in an ascii table
> for old 7 bit ascii stuff to be sure.
>
> >> thanks anoop! I guess those 02's are spaces then... Looks like
> >> most of the user lines from my old db are corrupted in this way as
> >> well. wierd.
>
> Not 02, but $20's, eg an ascii space char.
>
> >i would check the encoding being used vs. the encoding that was used
> > when it was initially created. it sounds like you were using a
> > wide_char encoding (eg. UTF-8) before and somehow has now reverted
> > back to latin-1 or some other single char encodings. i'm not an
> > expert on encodings etc.. know just enough to be dangerous. but if
> > this is database wide, (look at char/varchar/text fields and they
> > should all display this behavior), this is encoding related. on a
> > wide char encoding (say utf-8), the database reserves multiple bytes
> > per char not knowing what char it will need to save there. when you
> > tell mysql that it's not wide char, it will just show you what it has
> > - including the previously reserved bytes. it's odd that it's using
> > x20 to pad data tho.
again, the database wide funkyness with char fields suggests encodings not
lining up. after counting the chars in the output, it seems like it's
reserving 4 bytes per char (strlen("root")*4) (i know utf-8 causes mysql to
use three bytes per char), and because the later bytes (3 spaces per char)
were uninitialized under old mysql config under, presumably, a different
encoding - because in an encoding like UTF-8, ASCII chars are all 1 byte and
therefore "root" is the first 4 bytes. the current mysql just spits out x20
for the uninitialized reserved chars??
if(strlen(newFieldValue) == strlen(oldFieldValue) * 4) {
the tables were created using a variable wide chars encoding that supports
up to 4 bytes per chars. but now are under the 'default' (latin-1?) encoding.
}
else {
i dont' have enough encoding-fu to figure this out.
}
:)
> >
> >or something like that.
> >
> >> Thanks much for your help!
> >>
> >> matt
> >
> >--
> >
> >
> >anoop
> >[EMAIL PROTECTED]
>
> --
> Cheers, Gene
> People having trouble with vz bouncing email to me should add the word
> 'online' between the 'verizon', and the dot which bypasses vz's
> stupid bounce rules. I do use spamassassin too. :-)
> Yahoo.com and AOL/TW attorneys please note, additions to the above
> message by Gene Heskett are:
> Copyright 2006 by Maurice Eugene Heskett, all rights reserved.
--
anoop
[EMAIL PROTECTED]
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: VNC client/server combo doing VNC over HTTP
On Saturday 11 March 2006 02:02 am, Mark Fletcher wrote: > Hal Vaughan wrote: > >On Friday 10 March 2006 09:29, nullman wrote: > >>2 short infos to clarify : > >> > >>1. VNC over http doesn´t exist > >>2. Port-Numbers can be altered with any version > >> > >>Solution would be : ssh on Port 443 ... with that you can trick most > >>proxies with the "connect" method to use any proxy-capable ssh-client > >>(putty for example) > >>-> after ssh-connection is ok .. you can do vnc-over-ssh (simple > >>Port-forwarding) > > > >I couldn't get this to work in one of my situations, due to a nasty > >firewall. What I have found that seems to work is using stunnel to > >tunnel the VNC data through port 443 as HTTPS data, close to what is > >mentioned above. I'm still working on part of the solution, since I > >can't easily install stunnel on my clients Linux systems. When I'm all > >done, I'll post my results, since there has been very little on this > >list to directly apply to this -- at least on my case. > > > >Here's a link to stunnel: http://www.stunnel.org > > > >And here's a link to a tutorial about it, but it follows Windows, so > >you'll have to make some allowances and when they tell you to use > >ca.bat, it'll work best to download the file, extract the files that do > >the work, and convert them to Linux and run just those lines. You'll > >get some "directory does not exist" errors, but if you make the > >directory and re-run the program line, it'll work. At one point it'll > >complain about no index file, so do "echo 00 >index" and it'll fix it > >-- forgot what dir that is needed in, though. > > > >I'll have more detailed instructions later, when I've got all my stuff > >behaving at 100%. > > > >Hal > > Again thanks a lot for the suggestion, I'll try this too -- but I have a > possibly stupid question. What protocol will the gateway of my corporate > WAN think it is being asked to handle in this case? I don't think it > will allow any connections going out on VNC protocol, regardless of the > port number in use. HTTP / HTTPS is fine, not a lot else is... muhahaha... HTTPS, eh? excellent. try running ssh on port 443 at home and then try ssh-ing from work. the nice thing about HTTPS is that it's not a TLS type thing where you start off unencrypted and then do an encryption handshake. therefore, there shouldn't be *any* unencrypted data flowing back and fourth that the firewall can look at. the encrypted exchange is designed to stop man-in-the-middle. that takes the firewall out of the picture since it has nothing in the data flow that it can look at and go, "yes, it is indeed HTTPS". it's just relying on the port being 443. so any protocol should work as long as the port is 443. anoop. > > Am I just totally wrong on this? Or do I need to do something else to > disguise VNC packets as HTTP / HTTPS / something else a corporate > firewall can reasonably be expected to allow? > > Mark -- anoop [EMAIL PROTECTED]
Re: Turning off shell access
On Tuesday 14 March 2006 02:32 pm, Arnór Kristjánsson wrote: > How can I turn off shell access (through SSH) for certain users? "/bin/false" in the last field of /etc/passwd for that user should do the trick. -- anoop [EMAIL PROTECTED]
Re: PATH issues
On Friday 17 March 2006 11:03 am, [EMAIL PROTECTED] Davis wrote: > Hello. > > I have a problem with the PATH. I'm fairly familiar with the tradional > linux / unix roles for the various start up scripts, but I'm not getting > the behaviour that I expect. First of all, I'm running Ubuntu with Gnome. > I manually installed java in /opt/jdk1.5xxx The system had a free version > of java installed via APT. (I'm a bit of newbie when it comes to the > packaging system ). It appears that the APT process pretty much puts > everything into the typical, old school bins. These are the predefined > path. So, the APt installed free version of java is found on the > unmodified path, in /usr/local/bin or something. I need to have my version > of Java be found prior to that free one. ( I think if I understood the > packaging system there would be a more "debian" way of handling this, but > I'm not too hip to all of that ). > > My solution is to set the system wide path settings to include the bin > directory of my own java installation. PErhaps not the best way. Please > inform me of better solutions if you have them. I first tried adding this > PATH change to /etc/profile, and then to /etc/bash.bashrc. When I did > these things, the PATH changes were ONLY REFLECTED IN LOGIN SHELLS or, in > the case of the bashrc, terminal shells fired up from with in Gnome. The > problem is that when I try to launch the app that needs the new path > changes ( Eclipse won't run on the free java ?) from a Gnome launcher ( > desktop icon ), the PATH changes don't seem to be in effect. SO . . . why > don't the PATH changes effected by /etc/profile count when running > somehting from Gnome direclty? Perhaps this is a gnome issue? see the posting by "Jan T. Kim" with the subject "Re: login, path and ~/.profile" in this ML dated 09/30/05. what i have on my system for java is that: /usr/bin/java is a sym link to /etc/alternatives/java /etc/alternatives/java is a symlink to /usr/lib/j2sdk1.4-sun/bin/java if you want a different 'default' java implementation, change the /etc/alternatives/java symlink to point to a different executable. but then i've always just stuck with sun's jvm. haven't really juggled multiple jvms. > > > Chad -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Strange PPPoe problem
On Thursday 23 March 2006 01:13 pm, Jacob S wrote: > On Thu, 23 Mar 2006 12:27:26 -0500 > > Gene Heskett <[EMAIL PROTECTED]> wrote: > > On Thursday 23 March 2006 10:58, Jacob S wrote: > > >-BEGIN PGP SIGNED MESSAGE- > > >Hash: SHA1 > > > > > >Howdy list, > > > > > >I recently changed ISPs, away from static ips on a dsl line to a > > > single dynamic ip on Veriz*n's new Fi*S (fiber optic) service. The > > > new service uses PPPoe - not a problem, or so I thought - I have > > > PPPoe on my firewall. > > > > > >Now, I have used PPPoe from this very same firewall on a different > > >dsl line before and it worked great. But for some reason when I do > > >PPPoe for the new fiber line only http traffic works properly. When > > >downloading e-mail, everything is fine until it tries to download the > > >mail (I see it login, get the number of messages to download, and > > >then it tries to start downloading). At this point the e-mail just > > >hangs until it finally times out. It does not seem to be > > >port-related, as I have setup the e-mail server with port-forwarding > > >rules to allow me to download mail on non-standard ports and it > > >exhibits the same problem. And if I do PPPoe on the provided D-Link > > >router, instead of on my firewall, everything (including e-mail) > > >works great. > > > > Then I suggest you use it, as, provided you replace the d-link with a > > linksys, something like a BEFSX41, you'll also have a very good > > firewall for free AND it will all Just Work(TM). I spent 2 weeks > > trying to make rp's PPPoE for linux work but like you, way too many > > things just didn't work. > > > > The security of the d-link product has been questioned at length on > > the lists, and I can testify that the seimans speedstream product is > > likewise rather poor, it was owned and trashed here inside of 2 > > weeks, with outside config access supposedly denied from the WAN > > ports. > > > > My linksys has let someone by just far enough to make a log entry as > > they were being dropped by a combination of portsentry, tcpwrappers, > > and iptables, 3 times in 3 years, 2 of which came from known sources > > when one of vz dns servers was owned and attacked me. The third one > > came from a chinese address block and didn't get any farther that the > > log. For 3 years of 24/7/365 dsl service, I think thats very good > > security indeed. > > Thanks, but I'd rather keep playing with Linux to figure out why it's > not working i would too. ;) > than dump more money into the problem. My solution to this > point is using my firewall as the only computer connected to the D-Link > router. It works pretty well this way, but it means I'm stuck with > their "firewall" on the router, instead of having full control from my > Linux firewall. google PMTU to read about this in more detail, but it seriously sounds like icmp 3/4 packets are being dropped somewhere. if you setup your firewall to allow icmp packets of type 3/4 thru, you should be all set (well, you'd hope so anyway). a set of rules like so should do the trick: -A INPUT -p icmp --icmp-type fragmentation-needed -j ACCEPT -A OUTPUT -p icmp --icmp-type fragmentation-needed -j ACCEPT -A FORWARD -p icmp --icmp-type fragmentation-needed -j ACCEPT then, make sure you have the iputils-ping package installed (not the netkit-ping) and try: ping your.mail.host -c 1 -M do -s 1472 and you should get back an icmp reply saying what the mtu should be. subtract 28 from it and try pinging with that size and it should go thru. eg, if the reply says mtu = 1492, try: ping your.mail.host -c 1 -M do -s 1464 and it should go thru just fine. if you get a request timeout, that means that some routers are just dropping your packets without an icmp 3/4 message. keep reducing the size of your packet and see if you can get anything thru. read up on PMTU for possible solutions. there are ways to stop automatic PMTU discovery etc. hope it helps. anoop. > > Jacob -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Netlimiter liike tool
On Wednesday 22 March 2006 02:32 pm, Steve Lamb wrote: > Andreas Rippl wrote: > > Package: trickle > > First off it is really poor considering it needs to be run once per > application. This would require mass editing of init.d scripts. Secondly > trickle is very poor at what it does. > If you tell it to limit something to > 80kps and you have a 160kps line it will let that application burst to > 160kps for 1/2 second then cut it off completely for 1/2 second. can you define a finer grained timeslicing in trickle? > Not > exactly a trickle, more like a bipolar rainstorm. ;) -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Strange PPPoe problem
On Friday 24 March 2006 06:55 am, Jacob S wrote: > On Thu, 23 Mar 2006 14:35:20 -0600 > > anoop aryal <[EMAIL PROTECTED]> wrote: > > > > On Thursday 23 March 2006 10:58, Jacob S wrote: > > > > >-BEGIN PGP SIGNED MESSAGE- > > > > >Hash: SHA1 > > > > > > > > > >Howdy list, > > > > > > > > > >I recently changed ISPs, away from static ips on a dsl line to a > > > > > single dynamic ip on Veriz*n's new Fi*S (fiber optic) service. > > > > > The new service uses PPPoe - not a problem, or so I thought - I > > > > > have PPPoe on my firewall. > > > > > > > > > >Now, I have used PPPoe from this very same firewall on a > > > > >different dsl line before and it worked great. But for some > > > > >reason when I do PPPoe for the new fiber line only http traffic > > > > >works properly. When downloading e-mail, everything is fine > > > > >until it tries to download the mail (I see it login, get the > > > > >number of messages to download, and then it tries to start > > > > >downloading). At this point the e-mail just hangs until it > > > > >finally times out. It does not seem to be port-related, as I > > > > >have setup the e-mail server with port-forwarding rules to allow > > > > >me to download mail on non-standard ports and it exhibits the > > > > >same problem. And if I do PPPoe on the provided D-Link router, > > > > >instead of on my firewall, everything (including e-mail) works > > > > >great. > > > > > > > > > google PMTU to read about this in more detail, but it seriously > > sounds like icmp 3/4 packets are being dropped somewhere. if you > > setup your firewall to allow icmp packets of type 3/4 thru, you > > should be all set (well, you'd hope so anyway). a set of rules like > > so should do the trick: > > > > -A INPUT -p icmp --icmp-type fragmentation-needed -j ACCEPT > > -A OUTPUT -p icmp --icmp-type fragmentation-needed -j ACCEPT > > -A FORWARD -p icmp --icmp-type fragmentation-needed -j ACCEPT > > > > then, make sure you have the iputils-ping package installed (not the > > netkit-ping) and try: > > > > ping your.mail.host -c 1 -M do -s 1472 > > > > and you should get back an icmp reply saying what the mtu should be. > > subtract 28 from it and try pinging with that size and it should go > > thru. eg, if the reply says mtu = 1492, try: > > > > ping your.mail.host -c 1 -M do -s 1464 > > > > and it should go thru just fine. if you get a request timeout, that > > means that some routers are just dropping your packets without an > > icmp 3/4 message. keep reducing the size of your packet and see if > > you can get anything thru. read up on PMTU for possible solutions. > > there are ways to stop automatic PMTU discovery etc. > > Ok, things are getting stranger here. > > I ran the iptables rules you suggested and here's the ping results: > > # ping longbow.arroway.com -c 1 -M do -s 1472 > PING longbow.arroway.com (66.252.129.166) 1472(1500) bytes of data. > From pool-71-244-52-50.dllstx.fios.verizon.net (71.244.52.50) > icmp_seq=1 Frag needed and DF set (mtu = 1492) > > --- longbow.arroway.com ping statistics --- > 0 packets transmitted, 0 received, +1 errors > > # ping longbow.arroway.com -c 1 -M do -s 1464 > PING longbow.arroway.com (66.252.129.166) 1464(1492) bytes of data. > 1472 bytes from longbow.arroway.com (66.252.129.166): icmp_seq=1 ttl=49 > time=163 ms > > --- longbow.arroway.com ping statistics --- > 1 packets transmitted, 1 received, 0% packet loss, time 0ms > rtt min/avg/max/mdev = 163.150/163.150/163.150/0.000 ms > > So then I added the line > pty "/usr/sbin/pppoe -I eth0 -T 80 -m 1464" no, the mtu is 1492. you use 1464 in the ping because your are specifying the ping payload. the total packet size ends up being 1464+28 = 1492. (which is what you had to begin with.) all is not lost, we know PMTU works on your end provided you leave the three iptable rules mentioned above. it seems like a firewall on the mail server is causing icmp 3/4s from reaching the mail server. we can't do much about the firewall. ie, the mail server replied with packets with size = 1500 (most likely) with DF set, the other endpoint of your DSL sent back an icmp 3/4 message back to the mail server to send smaller packets, the firewall protecting the mail server dropped it and the mail server never knew to send you packets of 1492 or less. there is a 'hack
Re: Netlimiter liike tool
On Friday 24 March 2006 05:09 pm, Chris Roddy wrote: > indeed you can. trickle has several options for adjusting the detection > window and smoothing behavior. > > honestly it seems to me that this train-wreck has resulted from a poorly > phrased question. "is there a netlimiter-like tool in debian" is already > proposing the solution. it seems to me that the question here is "how > can i adjust network bandwidth consumption to achieve foo?" > > for me it was qos. now i've got my line saturated with bt traffic and > external services cruise along responsively. ymmv. > apart from having to run trickle for each instance of the app (i think the OP wanted a sys-tray type app that would throttle every invocation of an app x without having to wrap the command line for the app x with trickle.), seems to me that trickle does exactly what one would want. any throttling is really timeslicing. it just seems simultaneous because the timeslices are small enough. so, set a small timeslice, smooth it out.. then setup shell aliases such that firefox="trickle -trickle_opts firefox" the problem isn't 100% solved. so don't start flaming me ;) for instance, if you give firefox 50% of the bandwidth, and you launch 3 instances, you're hosed. another approach may be to script this following up and have it run from cron every so many seconds: 1. use netstat to find what ports are in use and what program/user has it open. for every open port: 2. lookup the programs/users in a conf file to see if they should be throttled. 3. if the program/user needs to be throttled, write a tc rule for that port. 4. remove the old tc rules, apply the new tc rules. this should handle dynamic ports. and since netstat can give you the user associated with the port, you can even do this by user, by program, by port and what have you. that too doesn't really fit the OPs requirements of not having to 'write-your-own'. but most other people should be fine with writing scripts. this is unix after all. so if it doesn't interest the OP (and please don't flame me down to a crisp), i hope it interests others to continue this discussion. this is an interesting problem after all. anoop aryal. [EMAIL PROTECTED] > > cmr > > anoop aryal wrote: > > On Wednesday 22 March 2006 02:32 pm, Steve Lamb wrote: > >> Andreas Rippl wrote: > >>> Package: trickle > >> > >> First off it is really poor considering it needs to be run once per > >> application. This would require mass editing of init.d scripts. > >> Secondly trickle is very poor at what it does. > >> > >> If you tell it to limit something to > >> 80kps and you have a 160kps line it will let that application burst to > >> 160kps for 1/2 second then cut it off completely for 1/2 second. > > > > can you define a finer grained timeslicing in trickle? > > > >> Not > >> exactly a trickle, more like a bipolar rainstorm. ;) -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Netlimiter liike tool
On Friday 24 March 2006 11:07 pm, Steve Lamb wrote: > Tony Godshall wrote: > > Looking for things in open source that work EXACTLY like > > commercial tools in Windows is a recipe for failure. > > Uh, no, it's not. It's pretty much the foundation of where a lot of OS > software pulls people away from Commercial applications. I mean would you > call someone trolling if they were asking "Is there an office suite like > Microsoft Office" (OpenOffice.org), or "an email/scheduling client like > Outlook" (Evolution)? No. So sorry, it isn't a recipe for failure... it's not a recipe for failure for asking. but it certainly becomes one when you start abusing the people who come up with something close *as best as they can*. you yourself have suggested that OO.o is a replacement for MSOffice. and Evolution for MSOutlook. is openoffice.org *EXACTLY* like MS Office? is Evolution *EXACTLY* like Outlook? i can come up with a set of criteria where OO.o would not be as EXACT a replacement as you seem to think. now, does the person recommending OO.o deserve a flogging? more to the point, you suggested that OO.o is a replacement for MSOffice. should i start a flame fest with you? # apt-get install mind-reader && read-mind neural-network unreachable. > > > Go troll somewhere else. > > ..so please take your own advice and troll elsewhere. -- anoop [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Strange PPPoe problem
On Saturday 25 March 2006 09:24 pm, Jacob S wrote: [snip] > > > # ping longbow.arroway.com -c 1 -M do -s 1472 > > > PING longbow.arroway.com (66.252.129.166) 1472(1500) bytes of data. > > > From pool-71-244-52-50.dllstx.fios.verizon.net (71.244.52.50) > > > icmp_seq=1 Frag needed and DF set (mtu = 1492) > > > > > > --- longbow.arroway.com ping statistics --- > > > 0 packets transmitted, 0 received, +1 errors > > > > > > # ping longbow.arroway.com -c 1 -M do -s 1464 > > > PING longbow.arroway.com (66.252.129.166) 1464(1492) bytes of data. > > > 1472 bytes from longbow.arroway.com (66.252.129.166): icmp_seq=1 > > > ttl=49 time=163 ms > > > > > > --- longbow.arroway.com ping statistics --- > > > 1 packets transmitted, 1 received, 0% packet loss, time 0ms > > > rtt min/avg/max/mdev = 163.150/163.150/163.150/0.000 ms > > > > > > So then I added the line > > > pty "/usr/sbin/pppoe -I eth0 -T 80 -m 1464" > > > > no, the mtu is 1492. you use 1464 in the ping because your are > > specifying the ping payload. the total packet size ends up being > > 1464+28 = 1492. (which is what you had to begin with.) all is not > > lost, we know PMTU works on your end provided you leave the three > > iptable rules mentioned above. it seems like a firewall on the mail > > server is causing icmp 3/4s from reaching the mail server. we can't > > do much about the firewall. ie, the mail server replied with packets > > with size = 1500 (most likely) with DF set, the other endpoint of > > your DSL sent back an icmp 3/4 message back to the mail server to > > send smaller packets, the firewall protecting the mail server dropped > > it and the mail server never knew to send you packets of 1492 or less. > > > > there is a 'hack' that may fix this. try and put this in your ruleset > > in the *mangle table (assuming your dsl interface is ppp0): > > > > -A POSTROUTING -p tcp --tcp-flags SYN,RST SYN -o ppp0 -j TCPMSS > > --clamp-mss-to-pmtu > > > > or use this from the command line if you don't already have the > > *mangle section: > > > > iptables -t mangle -A POSTROUTING -p tcp --tcp-flags SYN,RST SYN -o > > ppp0 -j TCPMSS --clamp-mss-to-pmtu > > > > what this does is sets the mss of the initial SYN packet to pmtu - > > 40. the reciever (the mail server in this instance), then, uses the > > mss to calculate the reply packet size instead of the default of > > 1500. by doing this it would have created packets of the right size > > (1492) and therefore doesn't need PMTU to work. > > > > again, the above rule is in addition to the previous three. > > Very nice! That time it worked. E-mail downloaded sucessfully several > times now. :-) I'll add those 4 lines to my firewall script. > > The mail server I'm downloading from is also a Debian machine... where > I have root access (even though it's not my own server). Is there > anything I should (can?) change in it's iptables firewall so that other > users don't have to fuss with this kind of a problem? apply the first three rules on the mail server. drop the fourth rule from your (home?) linux box. if you apply the first three rules to the mail server, the whole thing should work without the fourth rule. the fourth rule is a hack to make it work when the firewall on the other end drops icmp 3/4 messages. or you can leave it in... shouldn't really hurt. full explanation follows: ok, what you want to do is basically repeat the ping process but from the opposite end. now, you want your mail server getting the icmp 3/4 replies from your DSL providers router. in the diagram below: A B -X-[cloud]- C --- D let's say A is a client on your internal network. B the linux box where you have the iptables, C is the firewall protecting your mail server. and D is your mail server. note that A and B can be the same machine (the first three rules will have removed the distinction for our purposes.) and C and D can also be on the same machine... X is the other point in your dsl line (some router owned by your DSL provider). any how, what needs to happen is that when you ping with a packet size of 1500 from D to B (remember to subtract 28; just do what you did before but from the opposite end), you should get an icmp 3/4 from X to C, and C needs to let it thru to D. apply the first three rules to let icmp 3/4 thru in C. and ping again from D and see if you get "frag needed" messages back to D. if you do get the messages, you're done. if not, you'll need the clamp-mss rule since someone else is the PMTU black hole and you can't do anything about it. if you can get the icmp reply thru from X to C to D, you can discard the "clamp-mss" rule in B. or leave it in. because the whole purpose was to let D know that the link between B and X is 1492. you can let that be known by B initiating the connection with a suggestion to D on how big the reply should be. Or, you can let D find out about the 1492 link between B and X via the normal process of PM
Re: Help!
On Thursday 08 September 2005 06:11 pm, Bruno Buys wrote: > Paul E Condon wrote: > >On Thu, Sep 08, 2005 at 04:27:01PM -0400, Carl Fink wrote: > >>On Thu, Sep 08, 2005 at 10:52:50AM -0600, Paul E Condon wrote: > >>>For some products, Stephen's position is simply silly. Consider, > >>>for example, a Boeing 747. Another example is a C compiler. > >>>There are many more. Too many to list, and more than I could > >>>possible even know about. In general, the world is larger and > >>>more complicated than any of us know. > >> > >>It's a fine goal, even if it isn't always practical. For an installer, > >> it should be not just a goal but a requirement. (Docs should be > >> required only for very unusual situations.) > > > >OP had successfully installed Debian when he complained about not knowing > >what to do next. What he needed was some knowledge about what to do next, > >not about the installation. Like someone who successfully takes off in a > >747 and is annoyed that the thing doesn't land automatically. At least > >for Debian, it is not a life-threatening error. And only in a very narrow > >view of the world is documentation not required. Documentation, or > > training is not required only for those human activities that are common > > to all cultures. Sex, for example. > > Hey, sex does have its tricks... sex does have documentation: kamasutra. i pity da fool who doesn't RTFM. :P -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Laptop choice to run debian
On Saturday 10 September 2005 05:41 am, Hugo Vanwoerkom wrote: > Brian Kimball wrote: > > Antonio Rodriguez wrote: > >>I am planning to buy a laptop for my self in the near future for > >>travels, stinks not having with you a debian system. Any good > >>recommendations by any chance? > > > > A good recommendation that I don't see very often is to check which > > machines linux laptop vendors choose to use. Obviously they are going > > to pick machines that are as compatible as possible with linux. See > > > > http://www.emperorlinux.com/ > > Good idea. Just 3: IBM, Dell, Sony. i've always used a toshiba. the one i have currently is a desktop replacement. satellite pro 6100. therefore probably not ideal for what the OP wanted it for - travel. but otherwise, everything works out of the box. supports dual display (nvidia card). has an *excellent* display (1400x1050 - an awkward ratio but it works). and they have support website for linux. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: lock ssh user in his home
On Wednesday 14 September 2005 01:13 pm, Leonardo Marques wrote: > Hey people, > > Someone know what can i do to lock ssh user in his home ? a court order for a house arrest? ;) seriously tho, you may find rssh useful. i've used it to limit users to ssh or sftp or scp or a combination of 'em. and it's fairly trivial to setup chroot jail for the user. i think it even comes with a script to do just that. anoop aryal [EMAIL PROTECTED] > > Thanks for attention, > []s -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: GRUB problem
On Tuesday 27 September 2005 07:05 pm, Jeremy Merritt wrote: > I have been having a problem getting my GRUB > bootloader to return on boot. No matter what I do, it > keeps going to XP. I consulted with other people on > this list and got some good input. But have run into a > dead end again. Can someone analyze these steps and > tell me what I'm doing wrong, or what I need to do to > get the GRUB menu to return the way it was when I > first installed Debian: > > Steps so far: > > 1. Boot up with live CD (Knoppix) > 2. Activate shell and go root > 3. Mount /dev/hda5 as /mnt/hda5 > 4. chroot /mnt/hda5 /bin/bash > 5. grub-install /dev/hda5 -- Reports successful > install but no results on bootup i can't recall exactly what but i've had problems doing exactly what you described. and since, i've adopted a habit of keeping a GRUB boot disk around. (it's fairly trivial to create a bootable floppy. only slightly tougher to create a bootable CD.) i then use grub shell, to do a 'setup' of the drive. actually, it's easier to do this (i actually have a DHCP/network boot setup to launch grub should i not find my GRUB disk): 1) boot using the floppy/CD so that you have a GRUB screen. 2) type 'c' (i think) to get to the grub command line. 3) type 'root(hdX)' 4) type 'menu /boot/grub/menu.lst' 5) and boot like normal. once you get to your shell prompt, do a grub-install there. i wrote the instructions from memory so it might be a little our of wack. like i said, i can't remember the decisions that made me arrive at this solution but what i described above has bailed me out consistantly. hope it helps, anoop. [EMAIL PROTECTED] > > fdisk -l information (hda5=Mandrake, hdb2=Debian, > system is booting from hda1=XP): > > Disk /dev/hda: 40.0 GB, 40020664320 bytes > 255 heads, 63 sectors/track, 4865 cylinders > Units = cylinders of 16065 * 512 = 8225280 bytes > > Device Boot Start End Blocks Id System > /dev/hda1 * 1 2433 19543041 7 HPFS/NTFS > /dev/hda2 2434 4865 19535040 5 Extended > /dev/hda5 2434 3197 6136798+ 83 Linux > /dev/hda6 3198 3337 1124518+ 82 Linux swap > /dev/hda7 3338 4865 12273628+ 83 Linux > > Disk /dev/hdb: 61.4 GB, 61492838400 bytes > 255 heads, 63 sectors/track, 7476 cylinders > Units = cylinders of 16065 * 512 = 8225280 bytes > > Device Boot Start End Blocks Id System > /dev/hdb1 1 3188 25607578+ c W95 FAT32 (LBA) > /dev/hdb2 * 3189 7298 33013575 83 Linux > /dev/hdb3 7299 7476 1429785 5 Extended > /dev/hdb5 7299 7476 1429753+ 82 Linux swap > > Disk /dev/hdd: 30.0 GB, 30020272128 bytes > 255 heads, 63 sectors/track, 3649 cylinders > Units = cylinders of 16065 * 512 = 8225280 bytes > > Device Boot Start End Blocks Id System > /dev/hdd1 * 1 3649 29310561 83 Linux > > > > __ > Yahoo! Mail - PC Magazine Editors' Choice 2005 > http://mail.yahoo.com -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Webserver mirroring or web cluster tools
On Friday 30 September 2005 05:11 pm, Radhika wrote: > Hi, > > We are going to host the webserver for our ISP and we need to install > high availability.So is there any free webcluster software`with more > reliability and good performance or free mirroring software to > configure this.I am planning to install this on debian linux. if you want active/passive type cluster (HA, not load-balanced), use 'heartbeat' and 'stonith' and use apache on the machines. if you can make do with non-realtime replication of your web files, use rsync. if you need realtime replication, look into the kernel module drbd. use module-assistant to help you compile/install it if compiling kernels is not your thing. drbd is a 'RAID over TCP/IP' and has been working quite well for me. for a web server, use apache of course. configure it identically on both machines. etc.. - anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Want to move from root LVM/LILO to LVM/Grub
On Friday 30 September 2005 06:03 pm, Jason Martens wrote: > Hey all, >I kind of got myself into a pickle. no wonder the pickle tasted funny :P -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: authinticating email with a windows domain
On Tuesday 04 October 2005 09:57 am, theal wrote: > I am trying to set up smtp (exim) to authenticate with a windows domain > controller over a wan and I would like some idea on how to proceed. Any > help would be very much appreciated. 'winbind' might be what you're looking for. at least as a place to start. anoop. > > Tony -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: serial port programming
On Friday 07 October 2005 04:29 am, Marc Brünink wrote: > Hi, > > I've an application which uses /dev/ttyS3 to read and write some data. > This software is kind of alpha. Sometimes it segfaults. So the serial > port is not closed correctly. If I restart the application after a > segfault I'm not able to open the port again. I've to reboot. Im > wondering if it's possible to reset the serial port somehow. I mean: If > I sent a SIGKILL it'll result in the same behavior. And I should be > able to sent a SIGKILL to an application which uses the serial port, > shouldn't I? when you say can't open the port, do you mean that you can't do an open() on it? or open() works but can't read write anything? if you can open() it but not read()/write(), the settings that you tcsetattr() is probably additive to the initial state. anyhoo.. a quick and dirty way is to run minicom on the port, tweak the settings (flow control, baud rate etc) for the port in minicom until you can see the output in minicom. once you do, run 1) stty -g --file=/dev/ttySX 2) save that string somewhere. 3) then you can do, stty --file=/dev/ttySX may not be the cleanest way to do it but at least it saves you reboots. anoop. > > regards > Marc
PyKDE
whatever happened to PyKDE? i see python-qt3 but no pykde. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: PyKDE
On Friday 07 October 2005 10:42 pm, Jules Dubois wrote: > On Friday 07 October 2005 12:23, anoop aryal <[EMAIL PROTECTED]> > > (<[EMAIL PROTECTED]>) wrote: > > whatever happened to PyKDE? i see python-qt3 but no pykde. > > The presence in Debian of 'python-qt3' might lead one to generalize on > package-name conventions and deduce that there might exist another package > named 'python-kde3'. indeed. but upon such a generalization one quickly realizes that 'aptitude install python-kde3 && aptitude show python-kde3' reveals that it is not a real package. and one is left wondering what the hell happened to it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: PyKDE
On Saturday 08 October 2005 10:35 pm, Jules Dubois wrote: [ship] > > Although I couldn't be bothered to try it before, I see python-kde3 is a > dependency package. `aptitude install python-kde3` tells me: > > E: Unable to correct problems, you have held broken packages. > E: Unable to correct dependencies, some packages cannot be installed > E: Unable to resolve some dependencies! > ... > > The following packages have unmet dependencies: > python-kde3: Depends: python2.3-kde3 but it is not installable that's weird. on mine, i don't get any dependency issues. it 'installs'. but there are no files associated with it. will have to chase it down some day. ruby/korundum is looking better every day. > > I'm sorry for the attitude. don't worry about it ;) anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gui network scan
On Sunday 09 October 2005 06:23 am, Rodney Richison wrote: > What gui program might be good for scanning a network to take an > inventory of machines and ip address's etc on a network? try nmapfe anoop. > > -- > Highest Regards, > > Rodney Richison > RCR Computing > http://www.rcrnet.net > 118 N. Broadway > Cleveland, OK 74020 > 918-358- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: MySQL 4.1
On Tuesday 11 October 2005 09:55 am, Sanjay Debian wrote: > On 10/11/05, Michael Satterwhite <[EMAIL PROTECTED]> wrote: > > I'd been running MySQL 4.0 on Debian, but need to upgrade to 4.1. What > > procedure would you recommend? What about supporting PHP / Python / ??? > > packages > > > > The 4.1 packages appear to be broken as the command > > apt-get install mysql-server-4.1 mysql-client-4.1 mysql-common-4.1 > > [snip] > > I upgraded from mysql 4.0 to 4.1 a while back, without any issues. Try > doing apt-get update and then install mysql. There is also a switch to > apt-get to install with all the dependencies. I don't remember now, but you > would be able to find out from manpages. > > Hope this helps. > > sanjay try using 'aptitude'. eg: aptitude install mysql-server-4.1 it should do a better job of resolving dependencies. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ways to read man pages
On Saturday 22 October 2005 07:14 pm, kamaraju kusumanchi wrote: > Hi >We all know how important it is to read man pages. I would like to > collect info on different ways to read the man pages. My favorites are > as follows > > 1) man pagename > 2) In konqueror, man:pagename > 3) In vim, Man pagename > > What about others? emacs M-x man > > thanks > raju > > -- > Kamaraju S Kusumanchi > http://www.people.cornell.edu/pages/kk288/ > http://malayamaarutham.blogspot.com/ anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ssh chroot scp and sftp
On Sunday 23 October 2005 06:13 am, Rakotomandimby Mihamina wrote: > Hi, > > I am running Debian on a production box, and I look for a Debian package > to install that would make easy to chroot an user that tries to > scp/sftp. > > What is available? Thank you :-) rssh might be worth looking into. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Do I need to reinstall the whole system?
On Monday 24 October 2005 03:28 pm, Philippe Grenard wrote: > Hi all! > > well, last time I did a dist-upgrade , the system hanged, and even the > "magic keys" didn't answered, so I had to reboot the system. > > After that, there was one package that i couldn't upgrade, nor install, nor > uninstall. (xserver-xorg to be precise) > > So after somme googling, I tried to edit > /var/lib/dpkg/status > /var/lib/aptitude/pkgstates > and remove the entries of xserver-xorg > I also removed /var/lib/dpkg/info/xserver-xorg.* (preinst, postinst, prerm, > postrm, and so on) > > but the system still won't install the package. It always failed during the > preinst script it seems : > > root # dpkg -i > /var/cache/apt/archives/xserver-xorg_6.8.2.dfsg.1-9_i386.deb (Lecture de la > base de données... 134036 fichiers et répertoires déjà installés.) > Dépaquetage de xserver-xorg (à partir > de .../xserver-xorg_6.8.2.dfsg.1-9_i386.deb) ... > dpkg : erreur de traitement > de /var/cache/apt/archives/xserver-xorg_6.8.2.dfsg.1-9_i386.deb > (--install) : le sous-processus pre-installation script a retourné une > erreur de sortie d'état 1 > Des erreurs ont été rencontrées pendant l'exécution : > /var/cache/apt/archives/xserver-xorg_6.8.2.dfsg.1-9_i386.deb > > ( it's in french, but basically, it says that there was an error during > dpkg during preinst script with an error or exit code 1) > > Well, the questions are : > - How to make the system believe that the package isn't (or is) install, in > order to remove and reinstall it properly > - Is there a way to "run the dpkg process" step by step by hand with only > some basic commands as cp, rm, mv, mkdir, to see where there is a > problem - any other idea? > - Do I need to reinstall the whole system? i would try: aptitude reinstall and see if it helps. anoop. > > I thank you all for any help! > I already tried a lot of google but i couldn't find a solution to this > problem. > I also don't mind to read a lot of useful stuff if anybody has some ideas. > > Thanks a lot for any useful link or idea!!!
Re: Problems setting up Samba+LDAP PDC in Debian Sarge
On Wednesday 26 October 2005 06:48 am, Chema wrote: > Dear list, > > I have been struggling to get working a PDC using Samba with LDAP > backend, in a fresh Debian Sarge install. > > 1. SeMachineAccountPrivilege > > I'm reading IDEALX's Linux Samba-OpenLDAP Howto as guidance. In my > last attempt, everything appeared to be fine until the very end, the > Integration test, when I added an admin user, got it on the "Domain > Admin" and then tried to grant such group the > SeMachineAccountPrivilege: > > dellj81:/# net -U root%MyUnixRootPass rpc rights grant 'CORENA\Domain > Admins' SeMachineAccountPrivilege > Failed to grant privileges for CORENA\Domain Admins > (NT_STATUS_ACCESS_DENIED) > > Seems I have some kind of account problem here, since I can't make this > to work using root nor Manager. > > The Howto states: > > < exist and used (uid=0). > > Such a user is created when initializing the directory whith the > smbldap-populate script. > > >From Samba 3.0.12, it is now possible for admin users to join computers > > to the domain without using the "root" account." > ... > In fact, the 'root' account is needed in the first place so that the > SeXXX privileges can be set.>> > > The smbldap-tools didn't setup any root/uid=0 account in LDAP: > > dellj81:/# slapcat | grep -i ^uid: > uid: Administrator > uid: nobody > uid: admin > uid: chema > dellj81:/# slapcat | grep -i uidnum > uidNumber: 1004 > uidNumber: 998 > uidNumber: 999 > uidNumber: 1002 > uidNumber: 1003 > > So maybee that's what I'm missing, or should a standard (/etc/passwd) > root suffice? > > 2. net getlocalsid > > Anyway, after fiddling around looking for clues, I found that I no > longer can get my local sid: > lookup your SID from LDAP, then try: $ net setlocalsid [snip] then make sure your groups are mapped. try: $ net groupmap list etc.. i can't remember the exact steps off the top of my head. but i hope this gets you going. if not, post back here and cc me and i'll try to find my notes. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] SATA vs. SCSI
On Thursday 27 October 2005 06:33 pm, Gene Heskett wrote: > On Thursday 27 October 2005 18:56, Roberto C. Sanchez wrote: > >I will hopefully soon be building a server to donate to my church to > >replace a used one that I donated earlier this year. My question is > >this: Is SATA or SCSI preferrable? > > > >I am shooting for top notch reliability. I understand that components > >will occasionally fail. However, I have always understood that SCSI was > >preferable to ATA. Now that SATA is in the mix, I am not sure if that > >is still true. I have not kept up with the latest and greatest in terms > >of technology developments in that area. > > Speaking as the retired CE of a tv station where it was always said that > scsi was best, I have problems with that statement. Our first foray into > using a scsi based commercial server resulted in its getting converted to > ata disks fairly rapidly as the scsi raid lost a drive at 2 week > intervals. A single big atapi/eide drive turned out to be faster, and a > heck of a lot more reliable. > > Ditto here at home, I gave up on scsi tape drives about 18 months ago and > bought a 200GB atapi/eide drive & setup amanda's virtual tapes on it. It > has so far, been about 100x more dependable than the scsi tape ever was. > > Now we've gotten into the video server scene, again with the recommended > terrabyte raid, scsi3-320 or some such based with a 1394B (800 > megabits/second=100 megabytes) interface to the servers, and again scsi is > being a problem child with an occasional stutter while playing and always > a missed first word as it starts. Put the same program file on a single > internal big atapi/eide drive and the performance is 100% reading while > writing so we put in 2 drives per server. Now our main concern is that we > don't seem to be able to find an rsync workalike that does both branches > of the apple filesystem so that we can fabricate a darned near realtime, > live, online, redundant backup in case one server chassis should upchuck > in the middle of a program playback. > > If anyone has a clue how we can simulate an rsync run between 2 dual g5 > servers at 5 minute intervals, we're all ears. try unison: http://www.cis.upenn.edu/~bcpierce/unison/faq.html use it to sync a semi large personal music repo and seems to work fine. put it in cron or something to do it every 5 minutes. if you want realtime replication, i'd suggest drbd http://www.drbd.org/ it's "RAID over TCP/IP" and has been working great for me for over a year now. it probably puts a slight drag on performance (depending how you configure it) but for us, we can take the performance hit in exchange for the high availability it offers. don't know if it works on OSX etc. but then this isn't really an OSX mailing list. > > >Tha machine will be acting as a terminal server and also housing all the > >user home directories and probably a few other services. > > > >I am wondering what the rest of the world, at least as far as those that > >read this list, think. > > > >-Roberto > > IDE raid, at least raid5. Why pay $100 or more per drive for the > priviledge of poorer drive lifetimes and performance by using scsi. > Its not worth the headaches when it absolutely has to work right the > first time everytime. > -- > Cheers, Gene > "There are four boxes to be used in defense of liberty: > soap, ballot, jury, and ammo. Please use in that order." > -Ed Howdershelt (Author) > 99.35% setiathome rank, not too shabby for a WV hillbilly > Yahoo.com and AOL/TW attorneys please note, additions to the above > message by Gene Heskett are: > Copyright 2005 by Maurice Eugene Heskett, all rights reserved. hope it helps. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
books/doc for programmers and sysadmin on a debian system
is anyone aware of books/doc about how best to: 1) manage source-code (not just in a CVS/subversion but, how do you lay it out so that managing multiple projects is easy) 2) do .deb packaging easily. 3) manage package repos easily. 4) etc.. i have read the individual docs/books (auto-tools, cvs/subversion, .deb packaging etc) but nothing that really ties it together for a really nice and efficient setup for a *programmer and sysadmin on a debian system*. any ideas? some best-practices thrown together by the debian-uber-gurus would probably make for a nice such book. i'm specifically interested in setups that cover: directory layout, subversion, .deb packaging for multiple projects, emacs, python, c++ any pointers greatly appreciated. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: books/doc for programmers and sysadmin on a debian system
On Tuesday 01 November 2005 04:24 am, Eric Lavarde wrote: > Hi, > > you might want to have a look at: > http://www.debian-administration.org/ it seems to be geared towards sysadmin more than programming. basically, i would love to learn about how other programmers using debian have their environment setup so that it is efficient to manage. maybe that's a website waiting to be setup. hmm. i wonder if others would be interested. anoop. > > Eric > > anoop aryal wrote: > > is anyone aware of books/doc about how best to: > > > > 1) manage source-code (not just in a CVS/subversion but, how do you lay > > it out so that managing multiple projects is easy) > > > > 2) do .deb packaging easily. > > > > 3) manage package repos easily. > > > > 4) etc.. > > > > i have read the individual docs/books (auto-tools, cvs/subversion, .deb > > packaging etc) but nothing that really ties it together for a really nice > > and efficient setup for a *programmer and sysadmin on a debian system*. > > any ideas? > > > > some best-practices thrown together by the debian-uber-gurus would > > probably make for a nice such book. > > > > i'm specifically interested in setups that cover: > > > > directory layout, subversion, .deb packaging for multiple projects, > > emacs, python, c++ > > > > any pointers greatly appreciated. > > > > anoop. > > -- > Gewalt ist die letzte Zuflucht der Inkompetenz. > Violence is the Last Resort of the Incompetent. > Gwalt jest ostatnem schronieniem niekompetencji. > La violence est le dernier refuge de l'incompetence. > ~ Isaac Asimov -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ipssend and RaidMan on IBM xSeries
On Tuesday 15 November 2005 08:26 am, TAC Forums wrote: > > Any idea or link someone (google was not helpful, nor was IBM websites) ? > > I called up IBM and they said the server was not supported under > Debian GNU/Linux. they only supported Red Hat and Suse linux. > > Any ideas why IBM does not like to support debian Gnu/Linux? I > personally think they are going to miss out big time if they don't > look at supporting the Debian user community. we use Debian GNU/Linux on 8 xSeries 3xx boxes. hopefully, IBM will someday officially support it. posting just for the record. > > Debian is going to become the strongest and will last beyond all the > other GNU/linux distros. anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: .bash_history
On Monday 21 November 2005 06:07 pm, Adam Hardy wrote: > Roberto C. Sanchez on 21/11/05 22:30, wrote: > > On Mon, Nov 21, 2005 at 10:00:33PM +, Adam Hardy wrote: > >>thanks for the tips about C-r Is there some sort of documentation on > >> this? I'd like to see how much history it keeps. It's a seperate program > >> from history, right? > > > >HISTSIZE > > The number of commands to remember in the command history > > (see HISTORY below). The default value is 500. > > > > [...] > > > >To find other matching entries in the history list, type > > Control-S or Control-R as appropriate. ctrl-s will suspend your term in some cases. ctrl-q is your friend. > > This will search backward or > > forward in the history for the next entry matching the search string > > typed so far. Any other key sequence bound to a readline command will > > terminate the search and execute that command. For instance, a > > newline will termi- nate the search and accept the line, thereby > > executing the command from the history list. > > OK thanks. by default the keybindings are from emacs -- helpful if you use emacs. can be changed to work like the keystrokes in vi. read the section 'READLINE' in 'man bash'. > > I assume that my .bash_history gets converted to binary at some point by > bash. I can see that bash is using .bash_history now, but it's not clear > why the file should become binary on my system. > > > Adam anoop. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Cups, Samba and the terror called Windows :)
On Wednesday 23 November 2005 09:43 am, Martijn Marsman wrote: > Whoops, the testparm ouput is wrong, heres a new one without errors.. > > /Centauri:/var/lib/samba/printers/test# testparm > Load smb config files from /etc/samba/smb.conf > Processing section "[drivers]" > Processing section "[printers]" > Processing section "[print$]" > Loaded services file OK. > Server role: ROLE_STANDALONE > Press enter to see a dump of your service definitions > > # Global parameters > [global] > workgroup = AFAB > server string = %h Samba Print server > security = SHARE > syslog = 0 > log file = /var/log/samba/log.%m > max log size = 1000 > socket options = TCP_NODELAY SO_SNDBUF=8192 SO_RCVBUF=8192 > printcap name = cups > dns proxy = No > panic action = /usr/share/samba/panic-action %d > printing = cups > print command = > lpq command = %p > lprm command = > > [drivers] > path = /var/lib/samba/printers/test > guest ok = Yes > > [printers] > comment = All Printers > path = /tmp > read only = No > guest ok = Yes > printable = Yes > browseable = No > > [print$] > comment = Printer Drivers > path = /var/lib/samba/printers/test > guest ok = Yes [print$] comment = Printer Driver Download Area path = /var/lib/samba/printers browseable = yes public = yes write list = root @users chown root.root /var/lib/samba/printers; chmod 777 /var/lib/samba/printers from then on, anytime you configure a printer in windows, if the drivers aren't on the server, windows uploads them. and the next time (maybe on a different computer), should be able to download them so installing a printer is a double click away. anoop. > Centauri:/var/lib/samba/printers/test# > > / > > Martijn Marsman wrote: > > Hi all, > > > > When setting up cups and Samba to make an anonymous printserver for a > > lot of clients and 20 printers > > i noticed this works fine :D > > > > only 1 problem, > > > > when adding a printer to windows, it asks about the driver. > > in my samba config a pointed out a directory where the drivers are, > > but i dont know what files to put in there, > > just ppd files? or the contents of official driver cd? and should a > > make that dir a samba share? > > > > i really need to auto install the drivers. > > > > > > btw, iam using Appsocket in cups, the printer + drivers work fine in > > cups :) > > > > here is my smb.conf > > > > /Centauri:/var/log/samba# testparm > > > > Load smb config files from /etc/samba/smb.conf > > Unknown parameter encountered: "disable spools" > > Ignoring unknown parameter "disable spools" > > Unknown parameter encountered: "show add printer wizzard" > > Ignoring unknown parameter "show add printer wizzard" > > Processing section "[printers]" > > Processing section "[print$]" > > Loaded services file OK. > > Server role: ROLE_STANDALONE > > Press enter to see a dump of your service definitions > > > > # Global parameters > > [global] > >workgroup = TEST > >server string = %h Samba Print server > >security = SHARE > >syslog = 0 > >log file = /var/log/samba/log.%m > >max log size = 1000 > >printcap name = cups > >dns proxy = No > >panic action = /usr/share/samba/panic-action %d > >printing = cups > >print command = > >lpq command = %p > >lprm command = > > > > [printers] > >comment = All Printers > >path = /tmp > >create mask = 0700 > >guest ok = Yes > >printable = Yes > >browseable = No > > > > [print$] > >comment = Printer Drivers > >path = /var/lib/samba/printers > >guest ok = Yes > > > > Centauri:/var/log/samba#/ > > > > Well i hope its possible :D > > > > thnx in advance! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
