[Linux-cluster] gfs2 blocking tasks
Digimer
lists at alteeve.ca
Sun Aug 19 20:24:05 UTC 2012
The config looks fine. Sorry, I should have specified syslog;
/var/log/messages.
On 08/19/2012 03:05 PM, Bart Verwilst wrote:
> <cluster name="kvm" config_version="13">
> <logging debug="on"/>
> <clusternodes>
> <clusternode name="vm01-test" nodeid="1">
> <fence>
> <method name="apc">
> <device name="apc01" port="1" action="off"/>
> <device name="apc02" port="1" action="off"/>
> <device name="apc01" port="1" action="on"/>
> <device name="apc02" port="1" action="on"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="vm02-test" nodeid="2">
> <fence>
> <method name="apc">
> <device name="apc01" port="8" action="off"/>
> <device name="apc02" port="8" action="off"/>
> <device name="apc01" port="8" action="on"/>
> <device name="apc02" port="8" action="on"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="vm03-test" nodeid="3">
> <fence>
> <method name="apc">
> <device name="apc01" port="2" action="off"/>
> <device name="apc02" port="2" action="off"/>
> <device name="apc01" port="2" action="on"/>
> <device name="apc02" port="2" action="on"/>
> </method>
> </fence>
> </clusternode>
> </clusternodes>
> <fencedevices>
> <fencedevice agent="fence_apc" ipaddr="apc01" secure="on"
> login="device" name="apc01" passwd="xxx"/>
> <fencedevice agent="fence_apc" ipaddr="apc02" secure="on"
> login="device" name="apc02" passwd="xxx"/>
> </fencedevices>
> <rm log_level="5">
> <failoverdomains>
> <failoverdomain name="any_node" nofailback="1" ordered="0"
> restricted="0"/>
> </failoverdomains>
> <vm domain="any_node" max_restarts="2" migrate="live"
> name="firewall" path="/etc/libvirt/qemu/" recovery="restart"
> restart_expire_time="600"/>
> <vm domain="any_node" max_restarts="2" migrate="live"
> name="zabbix" path="/etc/libvirt/qemu/" recovery="restart"
> restart_expire_time="600"/>
> </rm>
> <totem rrp_mode="none" secauth="off"/>
> <quorumd device="/dev/mapper/iscsi_cluster_quorum"></quorumd>
> </cluster>
>
> dlm_control.log*, fenced.log, corosync.log are either empty or just
> contain entries until Aug 15 or ( in case of fenced.log ) "Aug 18
> 23:41:46 fenced logging mode 3 syslog f 160 p 6 logfile p 7
> /var/log/cluster/fenced.log". I was expecting at least some log output,
> since everything seems to work fine. Maybe enabling more debugging might
> bring something up ( but debugging is already on @ cluster.conf )
>
> Kind regards,
>
> Bart
>
> Digimer schreef op 19.08.2012 17:45:
>> On 08/19/2012 05:52 AM, Bart Verwilst wrote:
>>> Hi,
>>>
>>> I have a 3-node cluster in testing which seem to work quite well ( cman,
>>> rgmanager, gfs2, etc ).
>>> On (only) one of my nodes, yesterday i noticed the message below in
>>> dmesg.
>>>
>>> I saw this 30 minutes after the facts. I could browse both my gfs2
>>> mounts, there was no fencing or anything on any node.
>>>
>>> Any idea what might have caused this, and then go away?
>>>
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240067] INFO: task
>>> kworker/1:0:3117 blocked for more than 120 seconds.
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240182] "echo 0 >
>>> /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240296] kworker/1:0 D
>>> ffff88032fc93900 0 3117 2 0x00000000
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240302] ffff8802bb4dfb30
>>> 0000000000000046 ffff88031e4744d0 ffff8802bb4dffd8
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240307] ffff8802bb4dffd8
>>> ffff8802bb4dffd8 ffff88031e5796f0 ffff88031e4744d0
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240311] 0000000000000286
>>> ffff88032ffbd0f8 ffff8802bb4dfbc8 0000000000000002
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240316] Call Trace:
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240334] [<ffffffffa0570290>]
>>> ? gfs2_glock_demote_wait+0x20/0x20 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240340] [<ffffffff81666c89>]
>>> schedule+0x29/0x70
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240349] [<ffffffffa057029e>]
>>> gfs2_glock_holder_wait+0xe/0x20 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240352] [<ffffffff81665400>]
>>> __wait_on_bit+0x60/0x90
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240361] [<ffffffffa0570290>]
>>> ? gfs2_glock_demote_wait+0x20/0x20 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240364] [<ffffffff816654ac>]
>>> out_of_line_wait_on_bit+0x7c/0x90
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240369] [<ffffffff81073400>]
>>> ? autoremove_wake_function+0x40/0x40
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240378] [<ffffffffa05713a7>]
>>> wait_on_holder+0x47/0x80 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240388] [<ffffffffa05741d8>]
>>> gfs2_glock_nq+0x328/0x450 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240399] [<ffffffffa058a8ca>]
>>> gfs2_check_blk_type+0x4a/0x150 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240410] [<ffffffffa058a8c1>]
>>> ? gfs2_check_blk_type+0x41/0x150 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240421] [<ffffffffa058ba0c>]
>>> gfs2_evict_inode+0x2cc/0x360 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240432] [<ffffffffa058b842>]
>>> ? gfs2_evict_inode+0x102/0x360 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240437] [<ffffffff811940c2>]
>>> evict+0xb2/0x1b0
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240440] [<ffffffff811942c9>]
>>> iput+0x109/0x210
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240448] [<ffffffffa0572fdc>]
>>> delete_work_func+0x5c/0x90 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240453] [<ffffffff8106d5fa>]
>>> process_one_work+0x12a/0x420
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240462] [<ffffffffa0572f80>]
>>> ? gfs2_holder_uninit+0x40/0x40 [gfs2]
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240465] [<ffffffff8106e19e>]
>>> worker_thread+0x12e/0x2f0
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240469] [<ffffffff8106e070>]
>>> ? manage_workers.isra.25+0x200/0x200
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240472] [<ffffffff81072e73>]
>>> kthread+0x93/0xa0
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240477] [<ffffffff816710a4>]
>>> kernel_thread_helper+0x4/0x10
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240480] [<ffffffff81072de0>]
>>> ? flush_kthread_worker+0x80/0x80
>>> Aug 19 00:10:01 vm02-test kernel: [282120.240484] [<ffffffff816710a0>]
>>> ? gs_change+0x13/0x13
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240061] INFO: task
>>> kworker/1:0:3117 blocked for more than 120 seconds.
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240175] "echo 0 >
>>> /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240289] kworker/1:0 D
>>> ffff88032fc93900 0 3117 2 0x00000000
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240294] ffff8802bb4dfb30
>>> 0000000000000046 ffff88031e4744d0 ffff8802bb4dffd8
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240299] ffff8802bb4dffd8
>>> ffff8802bb4dffd8 ffff88031e5796f0 ffff88031e4744d0
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240304] 0000000000000286
>>> ffff88032ffbd0f8 ffff8802bb4dfbc8 0000000000000002
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240309] Call Trace:
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240326] [<ffffffffa0570290>]
>>> ? gfs2_glock_demote_wait+0x20/0x20 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240332] [<ffffffff81666c89>]
>>> schedule+0x29/0x70
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240341] [<ffffffffa057029e>]
>>> gfs2_glock_holder_wait+0xe/0x20 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240345] [<ffffffff81665400>]
>>> __wait_on_bit+0x60/0x90
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240353] [<ffffffffa0570290>]
>>> ? gfs2_glock_demote_wait+0x20/0x20 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240357] [<ffffffff816654ac>]
>>> out_of_line_wait_on_bit+0x7c/0x90
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240362] [<ffffffff81073400>]
>>> ? autoremove_wake_function+0x40/0x40
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240371] [<ffffffffa05713a7>]
>>> wait_on_holder+0x47/0x80 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240380] [<ffffffffa05741d8>]
>>> gfs2_glock_nq+0x328/0x450 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240391] [<ffffffffa058a8ca>]
>>> gfs2_check_blk_type+0x4a/0x150 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240402] [<ffffffffa058a8c1>]
>>> ? gfs2_check_blk_type+0x41/0x150 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240413] [<ffffffffa058ba0c>]
>>> gfs2_evict_inode+0x2cc/0x360 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240424] [<ffffffffa058b842>]
>>> ? gfs2_evict_inode+0x102/0x360 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240429] [<ffffffff811940c2>]
>>> evict+0xb2/0x1b0
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240432] [<ffffffff811942c9>]
>>> iput+0x109/0x210
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240440] [<ffffffffa0572fdc>]
>>> delete_work_func+0x5c/0x90 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240445] [<ffffffff8106d5fa>]
>>> process_one_work+0x12a/0x420
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240454] [<ffffffffa0572f80>]
>>> ? gfs2_holder_uninit+0x40/0x40 [gfs2]
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240458] [<ffffffff8106e19e>]
>>> worker_thread+0x12e/0x2f0
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240462] [<ffffffff8106e070>]
>>> ? manage_workers.isra.25+0x200/0x200
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240465] [<ffffffff81072e73>]
>>> kthread+0x93/0xa0
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240469] [<ffffffff816710a4>]
>>> kernel_thread_helper+0x4/0x10
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240473] [<ffffffff81072de0>]
>>> ? flush_kthread_worker+0x80/0x80
>>> Aug 19 00:12:01 vm02-test kernel: [282240.240476] [<ffffffff816710a0>]
>>> ? gs_change+0x13/0x13
>>> <snip, goes on for a while>
>>>
>>> Kind regards,
>>>
>>> Bart
>>
>> I usually see this when DLM is blocked. DLM usually blocks on a failed
>> fence action.
>>
>> To clarify; this comes up on one of the three nodes only? On the node
>> with these messages, you shouldn't be able to look at the hung FS on the
>> effected node.
>>
>> Can you share you versions and cluster.conf please? Also, what is in the
>> logs in the three or four minutes before these messages start? Anything
>> interesting in the log files of the other nodes around the same time
>> period?
>>
>> digimer
>
--
Digimer
Papers and Projects: https://alteeve.ca
More information about the Linux-cluster
mailing list