== Comment: #0 - SUBRATA MODAK <Subrata Modak <subrata.ibm.com>> - 2011.05.10 05:51:37 == ---Problem Description--- This message is consistently produced during every system boot of RHEL5.7-Alpha: "rm: cannot remove `/var/run/subsys/rhsm': Is a directory" The sequence is: " ... Remounting root filesystem in read-write mode: [ OK ] Mounting local filesystems: [ OK ] Enabling local filesystem quotas: [ OK ] rm: cannot remove `/var/run/subsys/rhsm': Is a directory Enabling /etc/fstab swaps: [ OK ] ... " Regards-- Subrata ---uname output--- Linux 2.6.18-259.el5 #1 SMP Fri Apr 29 01:11:49 EDT 2011 ppc64 ppc64 ppc64 GNU/Linux Machine Type = 8231-E2C-10708FP ---Debugger--- A debugger is not configured ---Steps to Reproduce--- Mentioned Above ---Installation and Packaging Component Data--- Install disk info: N/A Install method: NIM Install ISO Information: RHEL5.7-Server-20110430.2-ppc-DVD.iso Userspace tool common name: File System Quota The userspace tool has the following bit modes: N/A Userspace rpm: N/A Userspace tool obtained from project website: na == Comment: #2 - ARUNA H. BALAKRISHNAIAH <arunabal.com> - 2011.05.16 11:54:50 == initscript patch Hi Subrata, Please test with the patch and share the results. rpm -qf /etc/rc.d/rc.sysinit initscripts-8.45.35-2.el5 Thanks, Aruna. == Comment: #3 - ARUNA H. BALAKRISHNAIAH <arunabal.com> - 2011.05.17 01:24:45 == Hi Subrata, I rebooted the machine with the patch , i couldnt find the error messages. Remounting root filesystem in read-write mode: [ OK ] Mounting local filesystems: [ OK ] Enabling local filesystem quotas: [ OK ] Enabling /etc/fstab swaps: [ OK ] Thanks, Aruna.
Created attachment 499271 [details] initscript patch
------- Comment From Subrata Modak subrata.ibm.com 2011-05-17 05:01 EDT------- Hi Aruna, This solves the issue. Thanks for the Patch. RH, The proposed patch solves the problem. Can you please include this patch inside RHEL5.7 ?
This has nothing to do with anaconda. The fact that there's even a patch to an initscript is a pretty good indicator of that.
------- Comment From Subrata Modak subrata.ibm.com 2011-05-17 09:57 EDT------- Correct, So, can you please let us know whether this patch is being looked to be included soon in RHEl5.7 ? Regards-- Subrata
That patch is very very wrong. You want to add a special case for rhsm like the other special cases, not blindly remove everything. Although /var/run/subsys/rhsm is a pretty bizarre directory. What's it putting there, and what's the lifecycle for that data?
This request was evaluated by Red Hat Product Management for inclusion in a Red Hat Enterprise Linux maintenance release. Product Management has requested further review of this request by Red Hat Engineering, for potential inclusion in a Red Hat Enterprise Linux Update release for currently deployed products. This request is not yet committed for inclusion in an Update release.
------- Comment From Subrata Modak subrata.ibm.com 2011-05-18 01:39 EDT------- (In reply to comment #10) > That patch is very very wrong. > > You want to add a special case for rhsm like the other special cases, not > blindly remove everything. > > Although /var/run/subsys/rhsm is a pretty bizarre directory. What's it putting > there, and what's the lifecycle for that data? We would request you to kindly provide a solution. Since this directory is containing nothing and the intention of the script is to remove the directory (which will never occur with 'rm -f'), we found this to be the optimum solution. Let us know how would you like to solve this ? Regards-- Subrata > > This request was evaluated by Red Hat Product Management for inclusion in a Red > Hat Enterprise Linux maintenance release. Product Management has requested > further review of this request by Red Hat Engineering, for potential > inclusion in a Red Hat Enterprise Linux Update release for currently deployed > products. This request is not yet committed for inclusion in an Update > release.
Created attachment 499535 [details] Corrected patch ------- Comment on attachment From arunabal.com 2011-05-18 03:42 EDT------- Have included a special case to check for rhsm, since it couldnt be included directly , because afile = /var/run/*, it could match only 1 directory below it, so its checked for subsys and there its checked for rhsm. Thanks, Aruna,
*** Bug 705730 has been marked as a duplicate of this bug. ***
------- Comment From shubgoya.com 2011-05-24 14:19 EDT------- I checked on couple of system X platforms and the message mentioned in bug description was appearing consistently during system boot of RHEL5.7Alpha. So this issue is reproducible in x systems too.
This patch still looks very wrong. WTF is rhsm doing with that directory?
To be more explicit: - /var/run/subsys isn't a defined directory used by anything else - why is rhsm using that? - what is it putting there? I shouldn't have ACKed this; it's not entirely clear that this is an initscripts bug.
------- Comment From arunabal.com 2011-06-02 01:30 EDT------- Can you please suggest the patch ?
This shouldn't be fixed in intiscripts at all. Harald, can we revert this patch? This is a bug in subscription manager: /usr/share/rhsm/subscription_manager/certlib.py: class ActionLock(Lock): PATH = '/var/run/subsys/rhsm/cert.pid' That should almost certainly be /var/run/rhsm/cert.pid - /var/run/rhsm is the directory that the package installs, for example.
Fixed on RHEL5.7 branch in 4c2b38d6e4c560b33fa6cfc5e626eae42fad4658, version 0.95.5.21
Fixed on master branch in 025de56c3f0192280560084c88660cad9b2ef788, version 0.96.2
tested on subscription-manager-firstboot-0.95.5.20-1.git.5.4445aa4.el5 in cirtlib.py, the path has been changed to: PATH = '/var/run/rhsm/cert.pid' as suggested. upon rebooting the machine, no error messages were found. Fixes should appear in 0.95.5.21 release. Marking VERIFIED.
Since this is fixed in subscription-manager I've reverted patch in initscripts
------- Comment From Subrata Modak subrata.ibm.com 2011-06-08 02:25 EDT------- Thanks. Will test on SNAP1 and post results soon. Regards-- Subrata
*** Bug 711826 has been marked as a duplicate of this bug. ***
------- Comment From Subrata Modak subrata.ibm.com 2011-06-09 07:25 EDT------- Hello RH, This is fixed in SS*2 release of RHEL5.7. Thanks very much. Regards-- Subrata
(In reply to comment #16) > This shouldn't be fixed in intiscripts at all. Harald, can we revert this > patch? > > This is a bug in subscription manager: > > /usr/share/rhsm/subscription_manager/certlib.py: > > class ActionLock(Lock): > > PATH = '/var/run/subsys/rhsm/cert.pid' > > That should almost certainly be /var/run/rhsm/cert.pid - /var/run/rhsm is the > directory that the package installs, for example. ah, tomatoes on my eyes... didn't see "subsys", my brain also thought of "/var/run/rhsm"
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHEA-2011-1078.html