在 2013-1-7，02:59，Vyacheslav Dubeyko <slava@xxxxxxxxxxx> 写道： > > On Jan 6, 2013, at 10:44 AM, Elmer Zhang wrote: > >> Hi, >> My server always crash and reboot when I start MySQLs which stores the data on NILFS2 filesystem. >> >> I found some vmcore, but they are so big and I don't know how to extract useful information from them. >> ~]# ll -h /var/crash/*/* >> -rw------- 1 root root 141M Dec 20 16:14 /var/crash/127.0.0.1-2012-12-20-16:14:03/vmcore >> -rw------- 1 root root 512M Jan 4 16:00 /var/crash/127.0.0.1-2013-01-04-15:59:20/vmcore >> -rw------- 1 root root 92M Jan 4 16:25 /var/crash/127.0.0.1-2013-01-04-16:25:35/vmcore >> -rw------- 1 root root 62M Jan 4 16:50 /var/crash/127.0.0.1-2013-01-04-16:50:02/vmcore >> -rw------- 1 root root 63M Jan 4 18:55 /var/crash/127.0.0.1-2013-01-04-18:55:43/vmcore >> -rw------- 1 root root 80M Jan 5 14:55 /var/crash/127.0.0.1-2013-01-05-14:55:15/vmcore >> -rw------- 1 root root 65M Jan 5 15:30 /var/crash/127.0.0.1-2013-01-05-15:30:35/vmcore >> -rw------- 1 root root 199M Jan 6 15:06 /var/crash/127.0.0.1-2013-01-06-15:06:19/vmcore >> > > Crashdumps can be useful, I think. As I can see, you can use crash utility for analyzing it (http://magazine.redhat.com/2007/08/15/a-quick-overview-of-linux-kernel-crash-dump-analysis/). I think that bt, log and ps internal commands of crash utility can be very interesting. Could you share details of its output? The output of crash: http://d.pr/n/V9ay bt: http://d.pr/n/fyAs log: http://d.pr/n/dYI0 ps: http://d.pr/n/KguM > >> And below is some snapshot of the screen, hope it's useful: >> http://d.pr/i/4OZR >> http://d.pr/i/10M3 >> http://d.pr/i/HnJn >> http://d.pr/i/HW6t >> > > I hope that these screenshots are a consequence of the initial issue. But maybe it is a some new issue. > >> My environment: >> kernel: 2.6.32-220.13.1.el6.x86_64 >> kmod-nilfs2-0.4.3-1.el6.x86_64 >> nilfs-utils-2.1.4-7.x86_64 >> Raid: raid10 with SEAGATE ST32000444SS(2TB) * 12 >> mount: /dev/sdb2 on /data0 type nilfs2 (rw,noatime,gcpid=9047) >> >> ~]# nilfs-tune -l /dev/sdb2 >> nilfs-tune 2.1.4 >> Filesystem volume name: (none) >> Filesystem UUID: dcfb7152-a342-48d0-a712-212a3062395e >> Filesystem magic number: 0x3434 >> Filesystem revision #: 2.0 >> Filesystem features: (none) >> Filesystem state: invalid or mounted,error >> Filesystem OS type: Linux >> Block size: 4096 >> Filesystem created: Mon Dec 3 13:56:51 2012 >> Last mount time: Sun Jan 6 15:41:34 2013 >> Last write time: Sun Jan 6 15:41:34 2013 >> Mount count: 25 >> Maximum mount count: 50 >> Reserve blocks uid: 0 (user root) >> Reserve blocks gid: 0 (group root) >> First inode: 11 >> Inode size: 128 >> DAT entry size: 32 >> Checkpoint size: 192 >> Segment usage size: 16 >> Number of segments: 1246464 >> Device size: 10456104173568 >> First data block: 1 >> # of blocks per segment: 2048 >> Reserved segments %: 5 >> Last checkpoint #: 6327388 >> Last block address: 256856089 >> Last sequence #: 3862375 >> Free blocks count: 598300672 >> Commit interval: 60 >> # of blks to create seg: 0 >> CRC seed: 0x3adfb6c3 >> CRC check sum: 0xca2168f3 >> CRC check data size: 0x00000118 >> > > Thank you for details. > > With the best regards, > Vyacheslav Dubeyko. > >> >> Thanks >> --- >> Elmer Zhang >> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in >> the body of a message to majordomo@xxxxxxxxxxxxxxx >> More majordomo info at http://vger.kernel.org/majordomo-info.html > --- Elmer Zhang -- To unsubscribe from this list: send the line "unsubscribe linux-nilfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
