|Summary:||rc.sysinit hangs on unclean boot|
|Product:||[Fedora] Fedora||Reporter:||Tony C <dr_caola>|
|Component:||initscripts||Assignee:||Bill Nottingham <notting>|
|Status:||CLOSED WORKSFORME||QA Contact:||Brock Organ <borgan>|
|Fixed In Version:||Doc Type:||Bug Fix|
|Doc Text:||Story Points:||---|
|Last Closed:||2007-01-20 02:33:06 UTC||Type:||---|
|oVirt Team:||---||RHEL 7.3 requirements from Atomic Host:|
Description Tony C 2006-03-22 14:54:50 UTC
Description of problem: After an unclean shutdown (e.g., power fail), /etc/rc.d/rc.sysinit hangs at the final âwaitâ statement (line 658). Booting can only be accomplished in this circumstance if âctrl-Câ is pressed and rc.sysinit is aborted. A clean shutdown and power-up or reboot results in a normal boot sequence. Version-Release number of selected component (if applicable): Initscripts-8.11.1-1 How reproducible: Every unclean shutdown. Steps to Reproduce: 1. FC4 system up-and-running 2. sync â then turn off or unplug (cringing the whole time) 3. Turn on system. Boot will hang in rc.sysinit as described Actual results: Boot hangs, with the last console message âEnabling swap space: OKâ Control-C will abort rc.sysinit, and boot will continue Expected results: Expect system to recover from power failure without manual intervention necessary. Additional info: I placed some âechosâ in rc.sysinit to confirm that boot was hanging at the âwaitâ statement. After liberally sprinkling some âwaitâ and âechoâ statements around the rc.sysinit script, it appears that the backgrounded statement that may be hanging is: action $"Mounting local filesystems: " mount -a -t nonfs,nfs4,smbfs,ncpfs,cifs,gfs -O no_netdev but I donât know enough about what is going on here to assert that with authority. For the record, I have util-linux-2.12p-9.14 installed.
Comment 1 Bill Nottingham 2006-03-22 19:16:23 UTC
If you enable sysrq, and hit alt-sysrq-t, what's it doing?
Comment 2 Tony C 2006-03-23 01:58:26 UTC
Created attachment 126522 [details] SysRq : Show State output during hang This is the alt-sysrq-T output during the hang.
Comment 3 Tony C 2006-03-23 02:03:09 UTC
I lost my comments when I attached the dmesg output, above! But the short story is that the attachment, above, is the alt-sysrq-T output produced in dmesg during the hang. I enabled sysrq by setting kernel.sysrq=1 in sysctl.conf, cut power, turned back on, and then when rc.sysinit hung, hit alt-sysrq-T, followed by ctrl-C to resume booting.
Comment 4 Bill Nottingham 2006-03-23 17:15:39 UTC
What does your /etc/fstab (and your stale /etc/mtab) look like?
Comment 5 Tony C 2006-03-23 18:26:59 UTC
Created attachment 126559 [details] /etc/fstab This is my /etc/fstab after a 'hung' boot is ctl-C'd to get through rc.sysinit.
Comment 6 Tony C 2006-03-23 18:28:12 UTC
Created attachment 126560 [details] /etc/mtab . . . and this is the corresponding /etc/mtab
Comment 7 Bill Nottingham 2006-03-23 18:43:10 UTC
Hm, I don't suppose removing the /proc/nfs/nfsd from fstab helps?
Comment 8 Tony C 2006-03-24 04:16:09 UTC
Will try removing /proc/nfs/nfsd from fstab, tomorrow. Thx! -T
Comment 9 Tony C 2006-03-26 21:25:21 UTC
Actually, it does help -- I commented out this line: none /proc/fs/nfsd nfsd defaults 0 0 and the system comes up after a power fail! I don't know exactly what that line does -- nor do I understand why /proc/fs/nfds would know or care about an unclean boot vs. a clean boot. Will I notice any problems if I just leave this out? Thanks!
Comment 10 Bill Nottingham 2006-03-27 19:48:11 UTC
Not sure why it would care, but removing it should not cause any problems.
Comment 11 Tony C 2006-03-27 21:35:31 UTC
When I was checking around the web to see if there were other reports of this happening, I noticed there are alot of reports that boot hangs after the "Enabling swap space: [OK]" message. Since that is how this error actually appears on-screen, I thought I would mention it, in case someone else is having this problem and doesn't realize it isn't related to swap space.
Comment 12 Christian Iseli 2007-01-20 00:17:12 UTC
This report targets the FC3 or FC4 products, which have now been EOL'd. Could you please check that it still applies to a current Fedora release, and either update the target product or close it ? Thanks.