#StackBounty: #linux #software-raid md_raid10 hangs – possible reasons?

Bounty: 50

I have been researching the possible cause for this sort of behaviour:

INFO: task md1_raid10:240 blocked for more than 120 seconds.
[123491.284102]       Not tainted 4.18.20 #1
[123491.284105] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[123491.284109] md1_raid10      D    0   240      2 0x80000000
[123491.284116] Call Trace:
[123491.284132]  ? __schedule+0x421/0x8c0
[123491.284138]  ? preempt_count_add+0x67/0xb0
[123491.284144]  schedule+0x3c/0x90
[123491.284162]  md_super_wait+0x69/0xa0 [md_mod]
[123491.284173]  ? remove_wait_queue+0x60/0x60
[123491.284184]  md_update_sb.part.61+0x4ad/0x900 [md_mod]
[123491.284197]  md_check_recovery+0x310/0x530 [md_mod]
[123491.284205]  raid10d+0x64/0x15a0 [raid10]
[123491.284211]  ? __schedule+0x429/0x8c0
[123491.284216]  ? schedule+0x46/0x90
[123491.284220]  ? schedule_timeout+0x1f7/0x370
[123491.284224]  ? preempt_count_add+0x67/0xb0
[123491.284229]  ? _raw_spin_lock_irqsave+0x25/0x50
[123491.284232]  ? preempt_count_add+0x67/0xb0
[123491.284243]  ? md_thread+0x120/0x160 [md_mod]
[123491.284252]  md_thread+0x120/0x160 [md_mod]
[123491.284258]  ? remove_wait_queue+0x60/0x60
[123491.284263]  kthread+0xf8/0x130
[123491.284272]  ? md_rdev_init+0xc0/0xc0 [md_mod]
[123491.284277]  ? kthread_create_worker_on_cpu+0x70/0x70
[123491.284281]  ret_from_fork+0x22/0x40

All I have found, including stackexchange points to some harsh server conditions under heavy system loads. In my case it occurs in a system which hosts 4HHS in Raid1+0 configuration, that are bright new, actually the system has been set up just yesterday and contains virtually no data. The hangup occurred during listing the home directory (which contains just 10 files) via an SMB client.

The disks on this system are encrypted with dm_crypt. No LVM is used.


Get this bounty!!!

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.