Hi,
I've done another reboot, took ages to get past starting the cups server (which it eventually did), but looking in var/log/messages, I saw this.
Aug 29 11:56:13 T7 kernel: oom-killer: gfp_mask=0x1d2 Aug 29 11:56:30 T7 kernel: DMA per-cpu: Aug 29 11:56:30 T7 kernel: cpu 0 hot: low 2, high 6, batch 1 Aug 29 11:56:30 T7 kernel: cpu 0 cold: low 0, high 2, batch 1 Aug 29 11:56:30 T7 kernel: Normal per-cpu: Aug 29 11:56:30 T7 kernel: cpu 0 hot: low 32, high 96, batch 16 Aug 29 11:56:30 T7 kernel: cpu 0 cold: low 0, high 32, batch 16 Aug 29 11:56:30 T7 kernel: HighMem per-cpu: empty Aug 29 11:56:30 T7 kernel: Aug 29 11:56:30 T7 kernel: Free pages: 880kB (0kB HighMem) Aug 29 11:56:30 T7 kernel: Active:96241 inactive:92748 dirty:0 writeback:0 unstable:0 free:220 slab:1861 mapped:188696 pagetables:838 Aug 29 11:56:30 T7 kernel: DMA free:16kB min:16kB low:32kB high:48kB active:6468kB inactive:6184kB present:16384kB Aug 29 11:56:30 T7 kernel: protections[]: 0 0 0 Aug 29 11:56:31 T7 kernel: Normal free:864kB min:864kB low:1728kB high:2592kB active:378496kB inactive:364808kB present:769984kB Aug 29 11:56:31 T7 kernel: protections[]: 0 0 0 Aug 29 11:56:31 T7 kernel: HighMem free:0kB min:128kB low:256kB high:384kB active:0kB inactive:0kB present:0kB Aug 29 11:56:31 T7 kernel: protections[]: 0 0 0 Aug 29 11:56:31 T7 kernel: DMA: 0*4kB 0*8kB 1*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 16kB Aug 29 11:56:31 T7 kernel: Normal: 0*4kB 0*8kB 0*16kB 5*32kB 11*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 864kB Aug 29 11:56:31 T7 kernel: HighMem: empty Aug 29 11:56:31 T7 kernel: Swap cache: add 394473, delete 394454, find 1204/1543, race 0+0 Aug 29 11:56:31 T7 kernel: Out of Memory: Killed process 3012 (nscd). Aug 29 11:56:31 T7 kernel: oom-killer: gfp_mask=0x1d2 Aug 29 11:56:32 T7 kernel: DMA per-cpu: Aug 29 11:56:32 T7 kernel: cpu 0 hot: low 2, high 6, batch 1 Aug 29 11:56:32 T7 kernel: cpu 0 cold: low 0, high 2, batch 1 Aug 29 11:56:32 T7 kernel: Normal per-cpu: Aug 29 11:56:32 T7 kernel: cpu 0 hot: low 32, high 96, batch 16 Aug 29 11:56:32 T7 kernel: cpu 0 cold: low 0, high 32, batch 16 Aug 29 11:56:32 T7 kernel: HighMem per-cpu: empty Aug 29 11:56:32 T7 kernel: Aug 29 11:56:32 T7 kernel: Free pages: 840kB (0kB HighMem) Aug 29 11:56:33 T7 kernel: Active:181407 inactive:7575 dirty:0 writeback:0 unstable:0 free:210 slab:1845 mapped:188772 pagetables:838 Aug 29 11:56:33 T7 kernel: DMA free:16kB min:16kB low:32kB high:48kB active:12180kB inactive:516kB present:16384kB Aug 29 11:56:33 T7 kernel: protections[]: 0 0 0 Aug 29 11:56:33 T7 kernel: Normal free:824kB min:864kB low:1728kB high:2592kB active:713448kB inactive:29784kB present:769984kB Aug 29 11:56:33 T7 kernel: protections[]: 0 0 0 Aug 29 11:56:33 T7 kernel: HighMem free:0kB min:128kB low:256kB high:384kB active:0kB inactive:0kB present:0kB Aug 29 11:56:33 T7 kernel: protections[]: 0 0 0 Aug 29 11:56:34 T7 kernel: DMA: 0*4kB 0*8kB 1*16kB 0*32kB 0*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 16kB Aug 29 11:56:34 T7 kernel: Normal: 0*4kB 3*8kB 0*16kB 3*32kB 11*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 824kB Aug 29 11:56:34 T7 kernel: HighMem: empty Aug 29 11:56:34 T7 kernel: Swap cache: add 394912, delete 394912, find 1254/1641, race 0+1 Aug 29 11:56:34 T7 kernel: Out of Memory: Killed process 3014 (nscd).
Other processes killed include such favourites as modprobe (and that's it - though it was killed 8 times!).
Is this related to cups and if not, where is it from as I think it should be in bugzilla rather than my desktop.
Something *is* up with cups though. If I try to use system-config- printer, the entire machine locks up, lots of HD thrashing and no action. I have to switch the machine off to get things working again.
Very annoying!
TTFN
Paul
On Sun, 29 Aug 2004 18:48:11 +0100, Paul paul@all-the-johnsons.co.uk wrote:
Is this related to cups and if not, where is it from as I think it should be in bugzilla rather than my desktop.
You want to find out if its cups? turn off the cups service and reboot again and examine the log for similar messages. Isolate what causes the problem if you can by turning off services one at a time.
As a counter-point I just booted into the 533smp kernel and I'm not seeing any problems with cups. I'm using a usb hp 3650 printer.
Something *is* up with cups though. If I try to use system-config- printer, the entire machine locks up, lots of HD thrashing and no action. I have to switch the machine off to get things working again.
Are you sure its cups? and not something hardware specific that cups just happens to set off? I can start the system-config-printer dialog without the cups service running.
-jef
Hi,
Something *is* up with cups though. If I try to use system-config- printer, the entire machine locks up, lots of HD thrashing and no action. I have to switch the machine off to get things working again.
Are you sure its cups? and not something hardware specific that cups just happens to set off? I can start the system-config-printer dialog without the cups service running.
cups worked yesterday on the 2.6.8-1.503 kernel and not today under the 532 kernel. I have done a cups shutdown and then restart. Massive lag, lots of HD thrashing then everything died and I needed a power reset to get back to normal.
On boot, took ages to get past the cups start up.
I've done an update to the 533 kernel and will see what that does.
I still suspect a problem with cups.
TTFN
Paul
Hi,
I've done an update to the 533 kernel and will see what that does.
I still suspect a problem with cups.
Going to 533 made no difference, big delays in boot up.
Looking harder this time, I can start system-config-printer, but when I try to edit the printer, that's when the disk thrashing starts. Any way of tracing if it's cups gone mad?
TTFN
Paul
On Mon, 30 Aug 2004 01:13:51 +0100, Paul paul@all-the-johnsons.co.uk wrote:
Hi, Looking harder this time, I can start system-config-printer, but when I try to edit the printer, that's when the disk thrashing starts. Any way of tracing if it's cups gone mad?
not to beat a dead horse... but i can edit my printer definition via system-config-printer without the cups service running. The cups service doesnt get started till i hit the apply button.
Try stopping the cups service... make sure everything looks okay with cups stopped then run s-c-printer and edit the entry... see if it thrashes before you hit apply... if it thrashes see if the cups process was restarted...
-jef
Hi,
Try stopping the cups service... make sure everything looks okay with cups stopped then run s-c-printer and edit the entry... see if it thrashes before you hit apply... if it thrashes see if the cups process was restarted...
Been there, seen it, done it, got the 2.1Mb strace generated file, kernel panic and lock up.
It's not right I tellls ye!
TTFN
Paul
Paul wrote:
Hi,
Try stopping the cups service... make sure everything looks okay with cups stopped then run s-c-printer and edit the entry... see if it thrashes before you hit apply... if it thrashes see if the cups process was restarted...
Been there, seen it, done it, got the 2.1Mb strace generated file, kernel panic and lock up.
It's not right I tellls ye!
TTFN
Paul
Do you have "rhgb quiet" at the end of the kernel line in grub.conf? If so, try booting with that removed. I was hanging at the cups startup during boot until I removed that.
Gerry
Hi,
Do you have "rhgb quiet" at the end of the kernel line in grub.conf?
Nope. I have it removed while using rawhide - so I can try and find where problems are.
If so, try booting with that removed. I was hanging at the cups startup during boot until I removed that.
Nope. It was still hanging with plenty of problems which seemed to be nscd related (probably symptomatic rather than actual cause).
TTFN
Paul
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On Monday 30 August 2004 14:16, Paul wrote:
Hi,
Do you have "rhgb quiet" at the end of the kernel line in grub.conf?
Nope. I have it removed while using rawhide - so I can try and find where problems are.
If so, try booting with that removed. I was hanging at the cups startup during boot until I removed that.
Nope. It was still hanging with plenty of problems which seemed to be nscd related (probably symptomatic rather than actual cause).
Condolences on this miserable problem.
I would be thinking about lsof -n before and during the thrashing. For example
# lsof -n >temp # (perform thrashing) # lsof -n >temp1 && diff temp temp1 | grep -v lsof
I just tried it here with just a pause and got no changes in open files, so there is some hope you might see interesting and relevant changes...
- -Andy
- -- Automatic actions for USB cameras, cardreaders, memory sticks, MP3 players http://warmcat.com/usbautocam
Hi,
# lsof -n >temp # (perform thrashing) # lsof -n >temp1 && diff temp temp1 | grep -v lsof
I just tried it here with just a pause and got no changes in open files, so there is some hope you might see interesting and relevant changes...
Can't see anything much in that. However, /var/log/cups/error_log is showing something of interest...
I [30/Aug/2004:14:29:39 +0100] Listening to 7f000001:631 I [30/Aug/2004:14:29:39 +0100] Loaded configuration file "/etc/cups/cupsd.conf" I [30/Aug/2004:14:29:39 +0100] Configured for up to 100 clients. I [30/Aug/2004:14:29:39 +0100] Allowing up to 100 client connections per host. I [30/Aug/2004:14:29:39 +0100] Full reload is required. W [30/Aug/2004:14:30:13 +0100] LoadDevices: Backend did not respond within 30 seconds!
(this was repeated quite a few times before this snippet)
Definately looks like a cups problem.
A previous error_log gave this
I [22/Aug/2004:05:14:22 +0100] Listening to 7f000001:631 I [22/Aug/2004:05:14:22 +0100] Loaded configuration file "/etc/cups/cupsd.conf" I [22/Aug/2004:05:14:22 +0100] Configured for up to 100 clients. I [22/Aug/2004:05:14:22 +0100] Allowing up to 100 client connections per host. I [22/Aug/2004:05:14:22 +0100] Full reload is required. I [22/Aug/2004:05:14:25 +0100] LoadPPDs: Read "/etc/cups/ppds.dat", 4227 PPDs... I [22/Aug/2004:05:14:27 +0100] LoadPPDs: Wrote "/etc/cups/ppds.dat", 4227 PPDs... I [22/Aug/2004:05:14:28 +0100] Full reload complete. I [22/Aug/2004:17:34:10 +0100] Job 66 queued on 'HP-DJ980cxi' by 'paul'. I [22/Aug/2004:17:34:10 +0100] Started filter /usr/lib/cups/filter/pstops (PID 8253) for job 66. I [22/Aug/2004:17:34:10 +0100] Started filter /usr/lib/cups/filter/foomatic-rip (PID 8254) for job 66. I [22/Aug/2004:17:34:10 +0100] Started backend /usr/lib/cups/backend/usb (PID 8255) for job 66. I [26/Aug/2004:09:21:51 +0100] Adding start banner page "none" to job 67. I [26/Aug/2004:09:21:51 +0100] Adding end banner page "none" to job 67. I [26/Aug/2004:09:21:51 +0100] Job 67 queued on 'HP-DJ980cxi' by 'bev'. I [26/Aug/2004:09:21:51 +0100] Started filter /usr/lib/cups/filter/pstops (PID 27300) for job 67. I [26/Aug/2004:09:21:51 +0100] Started filter /usr/lib/cups/filter/foomatic-rip (PID 27301) for job 67. I [26/Aug/2004:09:21:51 +0100] Started backend /usr/lib/cups/backend/usb (PID 27302) for job 67. I [28/Aug/2004:12:45:51 +0100] Scheduler shutting down normally.
TTFN
Paul