Kernel installed in wrong location

Fedora 27 x86_64. When DNF installs a new kernel, it isn't going into the right place (/boot) and is not detected by GRUB. Why not? For example, after installing the most recent new kernel, I see this. [root@machine ~]# ls /boot 7725dfc225d14958a625ddaaaea5962b config-4.14.11-300.fc27.x86_64 efi

Re: Kernel installed in wrong location

Rick Stevens wrote: >> Fedora 27 x86_64. When DNF installs a new kernel, it isn't going into >> the right place (/boot) and is not detected by GRUB. Why not? > Sure looks like the scriptlet didn't run. I think you should get a > report about the initramfs image and the system map having different

Re: Kernel installed in wrong location

Further, here is an extract of /var/log/messages for the attempted reinstall, surpressing the timestamp and dracut indicator. > dracut-046-8.git20180105.fc27 > Executing: /usr/bin/dracut > /boot/7725dfc225d14958a625ddaaaea5962b/4.15.9-300.fc27.x86_64/initrd > 4.15.9-300.fc27.x86_64 > dracut modu

Re: Kernel installed in wrong location

Further information from another attempt a few minutes ago. Could the broken pipe be the problem? What could cause that? /boot is at 52%. > # /usr/bin/dnf -y reinstall kernel*-4.15.9-300.fc27.x86_64 > Last metadata expiration check: 23:59:22 ago on Sun Mar 18 14:00:12 2018. > Dependencies resolved

Re: Kernel installed in wrong location

Ed Greshko wrote: >> [root@machine ~]# ls /boot/7725dfc225d14958a625ddaaaea5962b >> 0-rescue 4.13.10-200.fc26.x86_64 4.13.9-200.fc26.x86_64 >> 4.11.11-300.fc26.x86_64 4.13.11-200.fc26.x86_64 4.14.11-300.fc27.x86_64 >> 4.12.11-300.fc26.x86_64 4.13.12-200.fc26.x86_64 4.14.4-200.

Re: Kernel installed in wrong location

Ed Greshko wrote: >>> All of those directories are empty, yes? >> Most, but not all: >> >>> /boot/7725dfc225d14958a625ddaaaea5962b/4.15.8-300.fc27.x86_64: >>> initrd linux >>> /boot/7725dfc225d14958a625ddaaaea5962b/4.15.9-300.fc27.x86_64: >>> initrd linux > > > I get the feeling, based on the

Re: Kernel installed in wrong location

The following procedure did NOT fix my problem. Using RPM instead of DNF still did not install the new kernel in /boot. Note, the new kernel is not in /boot when I start. > # ls /boot > 7725dfc225d14958a625ddaaaea5962b > config-4.14.11-300.fc27.x86_64 > efi > extlinux > grub2 > initramfs-0-rescue

Re: Kernel installed in wrong location

Roger Heflin wrote: > I would run this and see where rpm thinks they should be since they > are not in /boot > rpm -qa --filesbypkg | grep -i vmlinuz Evidently, RPM thinks they should be in /boot (as I would expect): > $ rpm -qa --filesbypkg | grep -i vmlinuz > kernel-core /boot/.vmlinuz-4

Re: Kernel installed in wrong location

Ed Greshko wrote: >> It seems clear that some script or scriptlet in the package is not doing >> its job. But they seem pretty simple and I don't see any dependencies >> which would make them fail on this machine and not others. > That would seem to be a fair conclusion.  > > I think I would loo

Re: Kernel installed in wrong location

I think I found the problem. Comparing packages installed on this problem machine with others, I noticed that grubby was not installed. After installing it and then re-installing the kernel, I find the kernel in /boot where it belongs. Evidently, something isn't checking dependencies sufficiently.

Re: Kernel installed in wrong location

Greshko wrote: >> I think I found the problem. Comparing packages installed on this >> problem machine with others, I noticed that grubby was not >> installed. After installing it and then re-installing the kernel, >> I find the kernel in /boot where it belongs. >> >> Evidently, something isn't

Re: Kernel installed in wrong location

Ed Greshko wrote: > I have tested this in a VM and verified the failure. There is one > exception. If the /boot directory does not contain a directory with > the name of the "machine-id" then the upgrade will be successful. > > Will you be writing up the BZ? Bug 1559118. -- Dave Close

Consistent device naming

I manage a lab with more than a hundred machines, all doing the same things. Presently they are running Fedora 13 and are obviously sorely overdue for an upgrade. But doing an upgrade is likely to present several issues, one of which I'd like to discuss today. The machines come from several man

Re: Consistent device naming

On 05/13/2014 05:39 PM, Russell Miller wrote: >> I want to run the same software on all of these machines and >> having inconsistent names /between/ the machines makes that next to >> impossible. Using the new names means that my software has to learn >> all those different names and can't easily

Re: Consistent device naming

On 05/13/2014 06:53 PM, Bruno Wolff III wrote: > To exchange names, you need to use a temporary name and do three > renames. Pretty much like swapping to values using a temporary > variable. I can't find a way to do that with udev. Please enlighten me. > Another thing to consider is do you reall

Re: Consistent device naming

Bruno Wolff III wrote: >>> To exchange names, you need to use a temporary name and do three >>> renames. Pretty much like swapping to values using a temporary >>> variable. >> >> I can't find a way to do that with udev. Please enlighten me. > > I use "ip link set" to do that. I am not sure how you

Trying to ping from a subinterface.

On Fedora 20 x86_64. According to "man ping", ping should work from a subinterface specifying either the interface name or its address: > -I interface >interface is either an address, or an interface name. If inter‐ >face is an address, it sets source address to specified inter‐ >

Re: Trying to ping from a subinterface.

I wrote: > # ping -I eth3:sub1 172.17.30.1 > ping: SO_BINDTODEVICE: Invalid argument Rick Stevens responded: > "eth3:sub1" isn't an interface, it's an alias. The interface name is > the bit before the ":" (or "." in the case of a VLAN). > > If you were to do a "netstat -rn", you'd only see "eth3

Re: Trying to ping from a subinterface.

I wrote: > Thanks. And you are obviously correct, technically. But this command > worked as I expected on earlier versions of Fedora. We now have many > procedures with such commands embedded and were surprised to discover > they no longer work on F20. Why was it changed? I take it back. I was le

Can't upgrade F21 to F22

Output below. This has been repeated all day. Any thoughts? [root@dsusim ~]# cat /etc/fedora-release Fedora release 21 (Twenty One) [root@dsusim ~]# fedup --clean resetting bootloader config removing boot images removing downloaded packages removing miscellaneous files [root@dsusim ~]# yum u

Re: Fedora 22 - Change login screen resolution?

CS DBA wrote: > I'm playing around with Fedora 22 (KDE spin) on a macbook pro retina. I > can login and change the screen resolution and it stays permanent. > However the login screen is still tiny, anyone know how to change / > scale the resolution of the login screen? I've got some systems with

F23 PXE strange failure

I'm working with a new PXE configuration file which may certainly contain some errors. Usually, it terminates with a "pane dead". But after the latest modification, it terminated by going to the Dracut emergency shell. In itself, that might help me, even though thus far I haven't found my error

Re: PXE Fedora 23 with Anaconda 20?

I wrote: > I've been trying to get a new working PXE installation for F23 and it > isn't working. I just noticed what might be the problem. Although > I downloaded the vmlinuz and initrd.img from the F23 server > version archive, when the installation starts it shows F20 in the > upper-right and c

error setting up base repository

The subject message sometimes appears on the installation source entry from anaconda. But I have never been able to find a good way to learn just what the error is. Sure, sometimes it's obvious: the URL is wrong. But other times everything seems to be correct but still the error persists. I'm l

FW: Firefox 44 on Fedora 23 won't start

-Original Message- From: CLOSE Dave Sent: Thursday, February 18, 2016 06:17 PM Pacific Standard Time To: support-fire...@lists.mozilla.org Subject:Re: Firefox 44 on Fedora 23 won't start I wrote: > A fresh installation of Firefox 44 on a fresh installation o

Virtual desktops on F23 server edition

I don't seem to have any virtual desktops available after a fresh installation of F23 server edition. What am I missing? -- Dave Close -- users mailing list users@lists.fedoraproject.org To unsubscribe or change subscription options: https://admin.fedoraproject.org/mailman/listinfo/users Fedora

Re: Virtual desktops on F23 server edition

On 02/24/16 02:45 PM, Joe Zeff wrote: > Virtual desktops are normally a part of a GUI, which isn't part of the > default installation for the server edition. Do you mean that you don't > have the alternate text consoles that should be available? Yes. Sorry for the terminology confusion. -- Dave

Re: Virtual desktops on F23 server edition

stan wrote: >> I don't seem to have any virtual desktops available after a fresh >> installation of F23 server edition. What am I missing? > Do you have the following files on your system? > /usr/lib/systemd/system-generators/systemd-getty-generator > /usr/lib/systemd/system/console-getty.service

Re: Virtual desktops on F23 server edition

I wrote: > My file is identical to the default except, > >NAutoVTs=6 # same as default >ReserveVT=6# same as default >LidSwitchIgnoreInhibited=yes Hit send to quickly. That last should be "LidSwitchIgnoreInhibited=no". -- Dave Close, Thales Avionics, Irvine California USA. ce

Re: Virtual desktops on F23 server edition

stan wrote: > Run systemctl -a -t service > and search the output for tty. It's less, so /tty will do it. There > should be a service running for each getty. # systemctl -a -t service | grep tty getty@tty1.service loadedinactive deadGetty on tty1 > If there i

Re: Virtual consoles (was desktops) on F23 server edition

Rick Stevens wrote: > tty2-6 will not be started unless you try to use them (e.g. the getty > on tty2 will only start if you use ALT-F2 from a console screen or > CTRL-ALT-F2 from the desktop). Once they're started, however, they > continue to run. I fully understand this. I now believe the probl

Re: Virtual consoles (was desktops) on F23 server edition

Rick Stevens wrote: >> So my vconsoles are properly configured and work correctly outside of X. >> While X is running, if I press CTRL-ALT-F2, for example, the F2 vconsole >> does not start and the X display remains. However, X is stuck; keyboard >> or mouse input does not work. Remote access via

Installing F23 WS on Dell Inspiron 17 (5758)

Using the F23 workstation net install ISO, installation is nearly impossible because the touchpad is not working and the selected buttons are not clearly indicated. The touchpad works in the BIOS. I'd be happy to solve this issue after installation if I could get that done. But using TAB etc to

Re: Installing F23 WS on Dell Inspiron 17 (5758)

Ted Roche wrote: > Could you plug in a mouse? Ah, that's too simple! I'm embarrassed. Of course that works. But I standby my complaints about how it doesn't work to do it using the keyboard. Because the selected buttons are not clearly indicated and are not arranged in a predictable sequence.

DNF is a time waster

If I run something like, "dnf -y install ", and is already installed, DNF gives a response like the following. Last metadata expiration check performed 0:00:02 ago on ... Package is already installed, skipping. ... (long delay, sometimes several minutes) ... Dependencies resolved.

Re: DNF is a time waster

Several folks have responded that they don't see the same behavior I do. I don't doubt that. Even for me, the behavior is erratic, sometimes very slow, sometimes not. The specific instance which prompted my note was an attempt to install from a local RPM file, not even from a repository. The ma

F23 dracut can't find disk

I have five machines which were fresh-installed with F23 back in February and all have been booted successfully a few times since. Today, booting of all of them fails in exactly the same way: dracut says it can't find the disk filesystems. The kernel boots as it should, and of course that comes

Re: F23 dracut can't find disk

I wrote: > I have five machines which were fresh-installed with F23 back in > February and all have been booted successfully a few times since. > Today, booting of all of them fails in exactly the same way: dracut > says it can't find the disk filesystems. The kernel boots as it > should, and of c

Re: F23 dracut can't find disk

On 04/07/16 05:56 PM, Rick Stevens wrote: > Glad to help. Did the rebuild of the ramdisk work? Are you running > the new kernel? Just curious as this may happen on the next update, > too. I hope to have time for that tomorrow. -- Dave Close -- users mailing list users@lists.fedoraproject.org To

Re: F23 dracut can't find disk

I wrote: > I have five machines which were fresh-installed with F23 back in > February and all have been booted successfully a few times since. > Today, booting of all of them fails in exactly the same way: dracut > says it can't find the disk filesystems. The kernel boots as it > should, and of c

Re: F23 dracut can't find disk

On 04/21/16 05:05 AM, Harald Hoyer wrote: > If you change hardware or want to use the initramfs on another > machine, it is easier to build a generic initramfs. > > The generic mode can be activated by: > *-N, --no-hostonly > Disable Host-Only mode > > or > > * hostonly=no >

F23 kickstart stalls

Trying to kickstart a Dell R630 with ten total Ethernet ports, I see this output. Starting installer, one moment... anaconda 23.19.10-1 for Fedora 23 started. * installation log files are stored in /tmp during the installation * shell is available on TTY2 * when reporting a bug add logs fro

F23 KDE stability issues

I'll be the first to admit that I don't use Fedora in a conventional way. But it suits me -- or it did until things started misbehaving a few weeks ago. Now I'm finally disgusted enough to write some notes and hope for good advice. I run F23 with the KDE desktop in what used to be called runlev

Re: F23 KDE stability issues

On 06/14/16 09:33 PM, Tim wrote: >> When I leave my desk, I lock the screen but I disable locking based on >> inactivity. As long as this works, I'm happy. > > Obvious question: Just a simple screen lock, or is there a screensaver > running (some of them were terrible at crashing X)? > > I'd try

Kickstart dependency resolution

I have a local mirror of Fedora 23, kept current nightly. In the updates mirror, there is a plasma-workspace-5.6.5-1.fc23.x86_64.rpm and a plasma-workspace-common-5.6.5-1.fc23.x86_64.rpm. There are no copies of the 5.6.4-1 RPMs. Nonetheless, when I try to install to a particular machine, kickst

Kickstart with a static interface

I want to kickstart a laptop with (of course) just one Ethernet interface. I can make it DHCP during the kickstart but I need it to end up as a static interface after installation. Is that possible? How? -- Dave Close -- users mailing list users@lists.fedoraproject.org To unsubscribe or change s

Re: Kickstart with a static interface

I wrote: > I want to kickstart a laptop with (of course) just one Ethernet > interface. I can make it DHCP during the kickstart but I need it to end > up as a static interface after installation. Is that possible? How? I should add that the static address I need after installation is private, no

Re: Kickstart with a static interface

I wrote: > I want to kickstart a laptop with (of course) just one Ethernet > interface. I can make it DHCP during the kickstart but I need it to end > up as a static interface after installation. Is that possible? How? > I should add that the static address I need after installation is > private,

Dell Inspiron 17 with Fedora

This is just a notice for those who may care. I installed Fedora 23 on a Dell Inspiron 17 laptop. If I use the latest kernel, 4.6.6-200, the screen flickers incessantly except when the touchpad is used to move the cursor. If I use a slightly older kernel, 4.5.7-202, everything works fine. -- D

Cleaning up after an aborted upgrade

I have a machine where I tried to do a system-upgrade from F23 to F24. The download portion worked fine but after the reboot, the upgrade process aborted. I'd like to know why, but first I need to clean up the mess so I can try it again. Right now I have 4782 total packages installed including

Re: Cleaning up after an aborted upgrade

Samuel Sieb answered my plea for help. Thanks! > First do "rpm --rebuilddb" to make sure the rpm database is consistent. > Run "systemd-cat --version" to make sure you're currently running 229. > If so, then do "rpm -e systemd-222-14". Then try the upgrade again. I was running 229 (fc24) as you

Re: Cleaning up after an aborted upgrade

Samuel Sieb wrote: >> It seems to be trying to put everything back to F23. "rpm >> -qa|sort" shows that almost all of the fc23 packages are paired >> with fc24 packages, both of which are installed. Is there any >> reason I can't just remove the redundant packages? >> > You probably have both fedo

YUM check (Re: Cleaning up after an aborted upgrade)

Samuel Sieb wrote: >> Is there any counterpart of the old "yum check" where I can verify >> everything is now ok? >> > I don't know what "yum check" did. What are you trying to verify? From the man page for YUM. > check Checks the local rpmdb and produces information on any >problems

Strange F24 upgrade

I have several machines I've recently upgraded from F23 to F24 using DNF. Daily upgrades of F24 packages on most of them are proceeding as expected. But this morning I found one that did a very peculiar "upgrade". Details below. # cat /etc/fedora-release Fedora release 24 (Twenty Four) # rpm -q

install.img

I need a copy of the install.img file for Fedora 20. It doesn't seem to be present in archives.fedoraproject.org or in the F20 netinstall ISO. Can anyone point me to a copy? -- Dave Close ___ users mailing list -- users@lists.fedoraproject.org

Re: install.img

>> I need a copy of the install.img file for Fedora 20. It doesn't seem to >> be present in archives.fedoraproject.org or in the F20 netinstall ISO. >> Can anyone point me to a copy? On 10/31/16 11:47, jd1008 wrote: > See http://archives.fedoraproject.org/pub/archive/fedora/linux/releases/ Thanks

Re: install.img

On 10/31/16 12:23, jd1008 wrote: > Good grief You wqant to be spoonfed??? > > see > http://mirrordenver.fdcservers.net/fedora/linux/releases/20/Fedora/x86_64/os/ Well, maybe I'm blind, but that looks the same as fedoraproject to me. Is there some reason install.img is invisible to me? -- Dav

Re: install.img

On 10/31/16 12:11, stan wrote: > What is an install.img file? That is, what is its function? Is it > possible that it is actually called initrd.img? I've never seen a clear description of each of the files involved in a PXE installation, but install.img is certainly one of them. It may be the

Re: install.img

On 10/31/16 13:47, Rick Stevens wrote: > Note there are two files each in the pxeboot and isolinux directory > trees, initrd.img and upgrade.img. There is no install.img. There was > a major rehash of the way the installers worked at F22 or > thereabouts, so that might explain it. Checking the rel

Re: install.img

On 11/02/16 06:05, Michael Schwendt wrote: > FWIW, I only remember such a file [install.img] from older releases > of Fedora up to Fedora 14. For F20 there has been a separate > squashfs.img. > > So, maybe tell which instructions you're trying to follow. 1. install.img is present in the F24 direc

Problems with kernel 4.7.10-100.fc23

I have several machines still running Fedora 23 but otherwise kept current with all posted updates. The latest kernel, 4.7.10-100.fc23, is causing me several troubles. 1. On these machines, I run iptables but not firewalld. The only reason I need either is to provide a NAT service. With the lat

Excessive dependency

After upgrading FC27 to FC28 (x86_64), Thunderbird became very unreliable and useful extensions stopped working. I decided to take the easy solution and just downgrade Thunderbird back to what I had with FC27. But for some reason, DNF thinks OpenOffice is dependent on Thunderbird, and deletes it! I

Re: Excessive dependency

Samuel Sieb wrote: >> After upgrading FC27 to FC28 (x86_64), Thunderbird became very >> unreliable and useful extensions stopped working. I decided to take the >> easy solution and just downgrade Thunderbird back to what I had with >> FC27. But for some reason, DNF thinks OpenOffice is dependent o

Re: Excessive dependency

stan wrote: > Run this as > # dnf -x libreoffice\* --best --allowerasing --enablerepo=fedora27 > --enablerepo=updates27 downgrade thunderbird-52.9.1-1.fc27.x86_64 > > and it should point to the dependency of libreoffice you want to > downgrade that causes it to be erased. > ... > It is probably

Re: Excessive dependency

Ed Greshko wrote: > You're running F28 and trying to downgrade to packages in F27. > Basically trying to run a Franken version of Fedora. > > That is not something that is supported. I'm only trying to run an older version of Thunderbird. It's the dependencies that are killing me. -- Da

Re: Excessive dependency

Samuel Sieb wrote: > If you want 52.9 for F28, then you can get it from here: > https://koji.fedoraproject.org/koji/buildinfo?buildID=1104472 > > Then use versionlock or excludes to stop it from upgrading. Thank you! That is exactly what I needed. Now I have to figure out how to re-enable the a

Using rsync to get the Fedora repositories

I have a script that uses rsync to pull the Fedora repositories nightly. If I run it manually, it works flawlessly. But if it runs via cron, I get an error on all the Fedora 28 repositories. I do not get the error on Fedora 27 or RPMFusion 27 or 28 repositories. The script sets timeout values to 36

Re: Using rsync to get the Fedora repositories

Rick Stevens wrote: > Running interactively but not under cron usually indicates that you've > forgotten that running under cron, you don't have the same environment > as you do in an interactive session. So, make sure your environment is > set correctly in the script (such as the PATH and such).

Re: Using rsync to get the Fedora repositories

Yesterday, I wrote: > I have a script that uses rsync to pull the Fedora repositories > nightly. If I run it manually, it works flawlessly. But if it runs > via cron, I get an error on all the Fedora 28 repositories. I do not > get the error on Fedora 27 or RPMFusion 27 or 28 repositories. The > s

Resilient network configuration

On every version of Fedora in the last ten years or so, the basic network configuration is stored in files under /etc/sysconfig. When the machine boots or I run "systemctl restart network" (or "service network restart"), those files determine how the network gets set-up. After that, I can change th

Re: Resilient network configuration

I wrote: > For a primary interface with a static address, I may be able to add > the information as IPADDR2 (etc) in ifcfg-eth0. But how do I do this > if the primary interface gets its address via DHCP? Samuel Sieb answered: > Install "nm-connection-editor". It gives you a lot more > capabilit

Re: Resilient network configuration

Samuel Sieb wrote: > I just tried it and it worked. I was able to add a static address to > my wifi connection while keeping the dhcp one. > "nm-connection-editor" has to be run from the command line, there is > no icon for it. The Gnome connection editor doesn't support doing > this. Perhaps t

F29 to F30 VMware upgrade

I have upgraded three VMware images previously running Fedora 29 to Fedora 30 using system-upgrade. All the images are running on the same ESC hardware and, so far as I know, use the same VMware foundation. The first worked flawlessly. Both the second and third failed in exactly the same way: they

F30 createrepo

In previous installations, /usr/bin/createrepo was a symbolic link to /usr/bin/createrepo_c. Now with createrepo_c-0.14.1-1.fc30.x86_64, that link is no longer created. This disrupts lots of scripts and will force changes to a lot of documentation, including Fedora's own online. -- Dave Cl

Re: F29 to F30 VMware upgrade

Chris Murphy wrote: > Not sure what would corrupt it but there is competition for LBA 0, > the MBR, in that there's a bootloader portion in the first ~440 bytes > and then a partition table from that point until the 512th byte. So > whenever something changes a partition or a boot flag (active bi

F30 Zabbix

On F30, zabbix-agent requires "zabbix = 4.0.7-2.fc30" but zabbix provides "zabbix = 1:4.0.7-2.fc30". The result is that DNF will not update either. -- Dave Close ___ users mailing list -- users@lists.fedoraproject.org To unsubscribe send an email

Re: F29 to F30 VMware upgrade

Chris Murphy wrote: > We'd need to look at LBA 0 on a broken system to do an autopsy. Once > it's fixed, the evidence of what stepped on it is wiped away. I have two more systems to upgrade, probably next week. If one of them fails in the same way, I'll capture the MBR and post here. --

Re-install Fedora without destroying user data

On a default Fedora installation with a sufficiently large disk, /home is a separate filesystem and should not contain any required system files. Thus it ought to be possible to completely re-install Fedora, the same version as was previously installed, without over-writing /home. I'd like to do th

Re: Re-install Fedora without destroying user data

suomi () wrote: > as far as possible I do the installation(s) using a kickstart file. > It is not possible, when you install from a Live CD. > My disk-partitionning in the kickstasrt file looks like: > > # System bootloader configuration > bootloader --location=mbr --boot-drive=nvme0n1 > # Parti

Re: Re-install Fedora without destroying user data

suomi () wrote: > as far as possible I do the installation(s) using a kickstart file. > It is not possible, when you install from a Live CD. > My disk-partitionning in the kickstasrt file looks like: > > # System bootloader configuration > bootloader --location=mbr --boot-drive=nvme0n1 > # Partiti

Re: Re-install Fedora without destroying user data

I wrote: > Well, not so fast. suomi's example did not involve LVM but LVM is > the default for Fedora. Trying to follow the example and adjusting > for LVM doesn't work for me. With these kickstart commands, > > bootloader --driveorder=sda --location=mbr --boot-drive=sda > clearpart --none --i

Two independent SSHD processes?

I'd like to run two SSH daemons on the same machine. They will listen on different ports and have other differences in their configuration. I want these two daemons to be independent of each other: either can run while the other is stopped. I need to do this for recent (systemd) machines and fo

Re: Two independent SSHD processes?

Ian Pilcher and others wrote: > There is a "PidFile" option in the config file that can be used to > control this. But if you read the SSHD man page and the /etc/rc.d/init.d/sshd script, you'll find that the PID is always recorded at /var/run/sshd.pid and the PidFile variable is only used for r

Re: Two independent SSHD processes?

Ed Greshko wrote: >> But if you read the SSHD man page and the /etc/rc.d/init.d/sshd script, >> you'll find that the PID is always recorded at /var/run/sshd.pid and the >> PidFile variable is only used for reporting status. > And if you tried it? > > [root@meimei run]# grep pid /etc/ssh/sshd_conf

Re: Two independent SSHD processes?

Ed Greshko wrote: > Well, I don't "seem" to be telling you. I'm actually showing you. > :-) :-) Yes, and thank you. But you didn't explicitly mention the version. In my original note, I said I'd only tried SysV. Before I make the effort to do it on systemd, I thought it would be worth asking for

Re: Two independent SSHD processes?

Ed Greshko wrote: > You probably will want to follow this procedure in F21 > https://access.redhat.com/solutions/1166283 Thanks. That works perfectly in F20. So we appear to have confirmed that the default installation is broken in SysV versions (mine are F13). The linked page points to ano

Anaconda illegal address

When anaconda complains about an "illegal IP address string passed to inet_aton", how can I discover what the bad string contains? So far as I can see, all the configured addresses are perfectly valid. Here's a copy of the traceback: . -- Dave Close, Thales Avionics, Irvine

Re: Anaconda illegal address

Chris Murphy wrote: >> When anaconda complains about an "illegal IP address string passed >> to inet_aton", how can I discover what the bad string contains? So >> far as I can see, all the configured addresses are perfectly valid. >> Here's a copy of the traceback: . > If you

Re: Anaconda illegal address

Chris Murphy wrote: > I mean, when there's a crash libreport should pop up and offer to file > the bug for you. You need to enter bugzilla credentials and then it > automatically files the bug details including uploading all files. > Before it does this, it checks if a bug has already been filed w

Re: Anaconda illegal address

I wrote: > I will now open two BZ reports: > > 1. Anaconda should fail with a useful message when a required address > is missing. My report has now been closed, marked as a duplicate of BZ 1178320. That means at least one developer thinks providing a useless error message is not a bug. I strongl

Re: Anaconda illegal address

Joe Zeff wrote: > Closing a bug as a duplicate isn't saying that it's not a bug. It means > that the bug's already been reported, and that it's a waste of resources > keeping two versions of the report open. But it does mean that when the referenced report has already been closed and marked NOT

reposync for two Fedora versions

I've been running reposync successfully for Fedora 20 for months. Now I'd like to repeat the process for Fedora 21 while continuing to take updates for F20 on the same machine. But it isn't working for me. /usr/bin/reposync -n -a x86_64 -r fedora21 -p /data/yum-mirrors/x86_64 Error setting up re

F20 kickstart can't select a disk

Trying to install Fedora 20 via kickstart on a x86_64 server. This has worked previously on several other identical servers. On this one, installation stops at the first screen with the error, "no disks selected". Watching more carefully, the message is first, "failed to save storage configurat

Re: F20 kickstart can't select a disk

Tim wrote: > You should probably start by posting your kickstart file. Someone may > spot a problem with it. Well, I did say that the same identical kickstart file worked properly on several other identical servers. And I truly mean "identical". However, all the details are now in

Changing boot options through kickstart

Documentation tells me that kickstart allows an --append switch to the bootloader command to add additional text to the GRUB2 boot line. I haven't experimented with this and it is unclear to me if the change goes only into /boot/grub2/grub.cfg or if it is also put into /etc/default/grub. But ad

Re: Changing boot options through kickstart

I asked: > I want to delete "rhgb" and "quiet" on the boot line, preferably in > both places [/etc/default/grub and /boot/grub2/grub.cfg]. Is there > any way to do that through kickstart or do I have to use a > post-install step? Joe Wulf replied: > I think your cleanest way would be a simple on

Netcat anomaly

I have a Fedora 20 machine which is receiving UDP broadcast packets at regular intervals on a strange high port. No program is presently listening for these packets. If I run, "tcpdump -i eth0 port 29531", I see each of the packets arriving just as I expect. Note, the packets are not empty and con

Re: Netcat anomaly

I wrote: > I have a Fedora 20 machine which is receiving UDP broadcast packets > at regular intervals on a strange high port. No program is presently > listening for these packets. If I run, "tcpdump -i eth0 port 29531", > I see each of the packets arriving just as I expect. Note, the > packets ar

Re: Netcat anomaly

Rick Stevens wrote: > Quite odd. Seems to work on my machine. In one window, I did a > > nc -lu 29531 > > and in another: > > echo 'Hello there!' | nc -u localhost 29531 > > and the first window saw the message. Try the same and see if it > works. Yes, that works for me. Further, this works when

Broadcasts and netmask

I have a Fedora 20 machine which is receiving UDP broadcast packets at regular intervals on a high port. No program is presently listening for these packets. If I run, "tcpdump -iany port 29531", I see each of the packets arriving just as I expect. But if I then use Ncat as a surrogate for an appl

  1   2   >