Subject: Re: [vserver] I can't kill a processus in a vserver, I've dmesg error
From: "John A. Sullivan III" <jsullivan@opensourcedevel.com>
Date: Wed, 23 Dec 2009 09:23:39 -0500

 Wed, 23 Dec 2009 09:23:39 -0500
On Wed, 2009-12-23 at 15:10 +0100, KLEIN Stéphane wrote:
> Hi,
> 
> In a vserver, I've a processus with D status :
> 
> $ vps aux | grep "my vs1"
> 
> stephane   572 40004 my vs1       0.0  0.0   4704  1420 ?        D
> 11:41   0:01 scp -r root@example.com:/home/foobar/www/ ./
> stephane   573 40004 my vs1       0.2  0.0      0     0 ?        H
> 11:41   0:15 [ssh]
> 
> I can't kill this process :(
> 
> In dmesg, I've this :
> 
> [2220677.412013] BUG: unable to handle kernel paging request at 00020004
> [2220677.412048] IP: [<c1087710>] find get page+0x40/0x81
> [2220677.412080] *pde = 00000000
> [2220677.412102] Oops: 0000 [#9] SMP
> [2220677.412126] last sysfs file: /sys/devices/virtual/block/md1/md/mismatch cnt
> [2220677.412155] Modules linked in: iptable raw xt comment xt recent
> xt policy ipt ULOG ipt REJECT ipt REDIRECT ipt NETMAP ipt MASQUERADE
> ipt LOG ipt ECN ipt ecn ipt CLUSTERIP ipt ah ipt addrtype nf nat tftp
> nf nat snmp basic nf nat sip nf nat pptp nf nat proto gre nf nat irc
> nf nat h323 nf nat ftp nf nat amanda ts kmp nf conntrack amanda
> nf conntrack tftp nf conntrack sip nf conntrack proto sctp
> nf conntrack pptp nf conntrack proto gre nf conntrack netlink
> nf conntrack netbios ns nf conntrack irc nf conntrack h323
> nf conntrack ftp xt tcpmss xt pkttype xt physdev xt owner xt NFQUEUE
> xt NFLOG nfnetlink log xt multiport xt MARK xt mark xt mac xt limit
> xt length xt iprange xt helper xt hashlimit xt DSCP xt dscp xt dccp
> xt conntrack xt CONNMARK xt connmark xt CLASSIFY xt tcpudp xt state
> iptable nat nf nat nf conntrack ipv4 nf defrag ipv4 nf conntrack
> iptable mangle nfnetlink iptable filter ip tables x tables dummy
> hwmon vid i2c viapro loop evdev snd pcm snd timer psmouse serio raw
> snd soundcore snd page alloc i2c i801 i2c core pcspkr processor button
> intel agp agpgart rng core ext3 jbd mbcache raid1 md mod sd mod
> ide pci generic piix ide core crc t10dif ata piix ata generic ahci
> libata scsi mod r8169 mii thermal fan thermal sys [last unloaded:
> scsi wait scan]
> [2220677.412710]
> [2220677.412731] Pid: 558, comm: scp Tainted: G      D
> (2.6.31.6-vs2.3.0.36.24-beng #1) 986LCD-M/mITX
> [2220677.412775] EIP: 0060:[<c1087710>] EFLAGS: 00210203 CPU: 0
> [2220677.412801] EIP is at find get page+0x40/0x81
> [2220677.412824] EAX: 0001ffff EBX: f6bc46c0 ECX: 00020000 EDX: 00000000
> [2220677.412850] ESI: f3df42f0 EDI: 00020004 EBP: f6bc4778 ESP: dfcefd24
> [2220677.412876]  DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> [2220677.412902] Process scp (pid: 558, ti=dfcee000 task=f64f8df0
> task.ti=dfcee000)
> [2220677.412942] Stack:
> [2220677.412961]  00020000 00200296 f65ab400 00000444 f6bc477c
> f6bc46c0 f6bc4778 00000444
> [2220677.412998] <0> f6bc4778 c10879ea f6bc46c0 00000000 ffffffff
> c1087c4b 000000d0 00000444
> [2220677.413053] <0> 00000000 f6bc46c0 00000000 00444000 00000000
> f7e843f1 00445000 00000000
> [2220677.413118] Call Trace:
> [2220677.413141]  [<c10879ea>] ? find lock page+0x10/0x43
> [2220677.413166]  [<c1087c4b>] ? grab cache page write begin+0x30/0x88
> [2220677.413219]  [<f7e843f1>] ? ext3 write begin+0x79/0x1d2 [ext3]
> [2220677.413247]  [<c10884b3>] ? generic file buffered write+0xe8/0x27f
> [2220677.413287]  [<c1088a8e>] ?   generic file aio write nolock+0x26c/0x294
> [2220677.413318]  [<c10329d5>] ? current fs time+0x13/0x15
> [2220677.413344]  [<c10892f9>] ? generic file aio write+0x60/0xb7
> [2220677.413388]  [<f7e800c8>] ? ext3 file write+0x19/0x85 [ext3]
> [2220677.413421]  [<c10ae891>] ? do sync write+0xc0/0x107
> [2220677.413447]  [<c112dd35>] ? rb insert color+0x51/0xb5
> [2220677.413475]  [<c1040ad2>] ? autoremove wake function+0x0/0x2d
> [2220677.413502]  [<c10ae57c>] ? fsnotify access+0x5a/0x61
> [2220677.413529]  [<c10fd638>] ? security file permission+0xc/0xd
> [2220677.413555]  [<c10ae7d1>] ? do sync write+0x0/0x107
> [2220677.413580]  [<c10af1c5>] ? vfs write+0x7e/0xd6
> [2220677.413604]  [<c10af2b5>] ? sys write+0x3c/0x63
> [2220677.413630]  [<c10030e9>] ? syscall call+0x7/0xb
> [2220677.413653] Code: 0c 8b 44 24 10 e8 cd 59 0a 00 85 c0 89 c6 74 44
> 8b 08 b8 ff ff ff ff f6 c1 01 0f 45 c8 8d 41 ff 83 f8 fd 89 0c 24 77
> d5 8d 79 04 <8b> 59 04 89 7c 24 04 85 db 74 c7 8d 6b 01 89 d8 8b 7c 24
> 04 f0
> [2220677.413799] EIP: [<c1087710>] find get page+0x40/0x81 SS:ESP 0068:dfcefd24
> [2220677.413831] CR2: 0000000000020004
> [2220677.414173] ---[ end trace 90eddfda8e996248 ]---
> 
> I don't know if this information is useful for you.
> 
> Question : I can't kill this process, I need to reboot the server completely ?
> 
> Thanks for your help,
> Stephane
May I ask what kernel you are using? We hit a similar problem with
2.6.28 and 2.6.29.  Apparently it is a kernel bug fixed in subsequent
releases, i.e., if it is the same problem we experienced - John
-- 
John A. Sullivan III
Open Source Development Corporation
+1 207-985-7880
jsullivan@opensourcedevel.com

http://www.spiritualoutreach.com
Making Christianity intelligible to secular society