Re: Unusual /var problem.
On Thu, Feb 13, 2003 at 06:45:22AM -0800, Frank Tanner III wrote: > That's what it is. But it's odd. I don't stash any > files in the /var directory, so why would it use up > all of the inodes? Just a guess; there was a bug in mgetty that caused /var/log to fill: * Tue Aug 13 2002 Nalin Dahyabhai <[EMAIL PROTECTED]> 1.1.28-7 - rotate mgetty.log.unknown and mgetty.log.callback (#68049) - don't logrotate already-rotated logs (#68422) If you are running mgetty, check the version. Regards, Bill Rugolsky -- Psyche-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/psyche-list
Re: RH 9 - new glibc
On Thu, Mar 27, 2003 at 11:20:42PM -0600, Doug B wrote:
> If I understand correctly, it seems the reason to jump a major version
> is because the new glibc breaks binary compatability.
>
> OK... that makes sense.
>
> Why then would RedHat introduce a new glibc into RH 8 that breaks
> compatabilty with some programs (wine comes to mind and apparently
> VMWare)? Shouldn't they have patched the glibc version that shipped
> (and worked) rather than break RH 8 with a new version?
Do you know for sure that it breaks things? I'm about to install
it today ...
$ rpm -qp --changelog glibc-2.3.2-4.80.src.rpm | head -14
* Wed Mar 19 2003 Jakub Jelinek <[EMAIL PROTECTED]> 2.3.2-4.80
- build as 8.0 errata
- turn compatibility only symbols which are referenced in libc.so's
relocations back into @@ symvers, as otherwise statically linked
RHL 8 apps segfault when doing NSS or iconv
- move __libc_fork, __libc_stack_end, __libc_wait and __libc_waitpid
from GLIBC_PRIVATE symver for the errata
- remove {INIT,PREINIT,FINI}_ARRAY support, since it requires
binutils changes
- don't build NPTL libraries
- strip libraries and binaries, recreate glibc-debug-static
subpackage
Regards,
Bill Rugolsky
--
Psyche-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/psyche-list
Re: What's the status of PDA secure remote access?
On Fri, Aug 08, 2003 at 10:02:00AM -0400, Michael Martinez wrote: > I'm looking for a secure PDA solution that allows me to open a secure, > remote command line login to my Linux servers. I have specific need to > be satisfied (listed below). I did some research into this, 7 months > ago, and found that no vendor offered an adequate solution ... but that > each of the major vendors were at that time working on "emerging" > technology that *would* satisfy the solution. > > Anyone know of a solution that satisfies the following criteria for the > remote login process: > > 1. point to point encryption (ssh) with some sort of ssh client on the > PDA and an ssh server on the server > > 2. supports terminal emulation and the vi editor (being able to open and > edit text files during the remote login) > > 3. wide area coverage. In other words, if my system administrator is > stuck in traffic 30 miles out, she can use her PDA to log in to her > server Sharp Zaurus, and possibly an IR keyboard from Pocketop (so one can type at a reasonable speed) will likely do what you need. I have a Zaurus SL-5500 with a WiFi card, as well as a serial cable for my cell phone. If I have WiFi access, I use that, otherwise, there's the painfully slow 14.4K modem connection, which I can use to connect to the office, or Verizon's ISP service -- horribly slow, but it works "anywhere." By installing/upgrading packages on the Zaurus, one can use any of ssh, IPsec, PPTP, etc. There are several different terminal emulators that provide tabbed windows, so managing multiple sessions is simple. Regards, Bill Rugolsky -- Psyche-list mailing list [EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/psyche-list
Read the RELEASE-NOTES (was Re: True type fonts in mozilla./evolution)
On Thu, Nov 07, 2002 at 01:21:08PM +0100, Michael Schwendt wrote: > You couldn't. Unless, of course, you would invent something which > doesn't permit scrolling down to the bottom of the RELEASE NOTES, > clicking a checkbox and proceeding without having read the several > hundred lines of text. But maybe you could hide some keyword in the > text and force users to enter it at the bottom. ;) Many games have required that you find some random fact in the documentation to unlock the game. I'd shoot a bit higher, and require actual evidence of understanding -- fail the test, and you are thrown back to the documentation and given a different random question. :-) The same might be said for "click-through" licenses. Regards, Bill Rugolsky
Re: serial console question
On Mon, Nov 25, 2002 at 01:12:27PM -0500, Neal D. Becker wrote: > I'm wondering how to setup for serial console. I am setting up a > small cluster and want to avoid the expense of a kvm switch. > > Reading in the kernel serial console doc, I get the impression that I > don't need to setup anything? If there is no video card found I will > automatically get a serial console, configure 9600N8. Is this > correct, or do I need to edit some config file? The following should get you started. In /etc/grub.conf, make the first two lines: serial --unit=0 --speed=9600 terminal --timeout=10 --silent serial console [The timeout value is in seconds, the silent indicates that you don't want the "Press any key" prompt to pull up the GRUB menu.] If using LILO, the correct line will be serial=0,9600n8 Append the following to your kernel command line: console=tty0 console=ttyS0,9600 [In LILO, append="console=tty0 console=ttyS0,9600"] This will give you a vga console [if present], and a serial console. If there is no VGA device, then you can eliminate the first part. You will need to run a getty on the console, so in /etc/inittab, add the following: # run a getty on COM1 (serial console) co:12345:respawn:/sbin/agetty 9600 ttyS0 You can comment out the virtual console mingetty(s). I find 9600 annoyingly slow, but most hardware comes configured at that speed. Use 115200 throughout if your serial link can handle it. BTW, if you have not already done so, you probably want to turn on Magic SysRQ in /etc/sysctl.conf: # Enable the magic-sysrq key kernel.sysrq = 1 Then read Documentation/sysrq.txt in the kernel tree, and figure out how to send a BRK from your terminal software. Regards, Bill Rugolsky -- Psyche-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/psyche-list
Re: iptables logging
On Mon, Dec 02, 2002 at 11:34:54AM +0930, Iain Buchanan wrote: > I've got iptables to log what it drops with various options, but its > filling up my log files (theres a lot of traffic at work). Can I > somehow get it to log to a file other than /var/log/messages? Look into the ULOG target; you run ulogd, and you can slice-n-dice the messages however you please. ftp://ftp.netfilter.org/pub/ulogd/ Regards, Bill Rugolsky -- Psyche-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/psyche-list
Re: Kernel panic with ext3 root set to writeback
On Fri, Dec 20, 2002 at 09:12:26AM -0800, Tom Ball wrote: > title Red Hat Linux (2.4.18-18.8.0) - Home > root (hd0,5) > kernel /boot/vmlinuz-2.4.18-18.8.0 ro root=LABEL=/ rootflags=mode=writeback > initrd /boot/initrd-2.4.18-18.8.0.img > > As someone previously suggested, I also tried adding this information to > /etc/fstab, with the same panic. That's supposed to be rootflags=data=writeback Regards, Bill Rugolsky -- Psyche-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/psyche-list
Re: Much Slower than 7.2?
On Mon, Dec 23, 2002 at 11:20:25AM -0800, Tom Ball wrote: > I also found 8.0 much slower for my smaller builds. One thing that's > different in 8.0 is the ext3 filesystem, which I converted to when > upgrading. By default it runs in ordered data mode, which causes writes > to disk to happen much more frequently than on ext2 filesystems. Yes, particularly on /tmp. Traditionally, ext2 was so fast that the performance of /tmp was not an issue, but journalling changes that. I've been using tmpfs on my systems without problems. There are a few corner cases that have been cleaned up recently (see lkml), but in practice it has not been a problem. tmpfs need never write to swap if there is no memory pressure; it is a useful complement to the use of journalling filesystems for truly persistent data. Remember that use of meta-data only journalling poses a potential information leakage (i.e., security) problem on a multiuser system, especially when used on /tmp ... To use tmpfs, one can just add a line like the following to /etc/fstab: /tmp/tmptmpfs defaults0 0 See mount(8) in the man pages for details of the available options: Mount options for tmpfs The following parameters accept a suffix k, m or g for Ki, Mi, Gi (binary kilo, mega and giga) and can be changed on remount. size=nbytes Override default size of the filesystem. The size is given in bytes, and rounded down to entire pages. The default is half of the memory. nr_blocks= Set number of blocks. nr_inodes= Set number of inodes. mode= Set initial permissions of the root directory. Regards, Bill Rugolsky -- Psyche-list mailing list [EMAIL PROTECTED] https://listman.redhat.com/mailman/listinfo/psyche-list
Re: Rsync @ Red Hat
On Mon, Sep 30, 2002 at 02:54:28PM -0400, Chip Turner wrote:
> You should be able to use 'wget' on the URLs from the Instant ISO page
> and resume your downloads. Using your browser to download is highly
> unsafe; a single crash can lose the entire download. wget should work
> fine, but be sure to put the URL inside of single quotes (') to ensure
> that shell escaping does not damage the URL.
Alternatively, copy all of the URL's to a file (/tmp/urls), and use
wget -i /tmp/urls
Regards,
Bill Rugolsky
