Topic: raid resyncs weekly but status is UU for md1 and md2

[mike_mattos]

Had to replace a system board, and ever since mdstat fires every Sunday, but the mdstat  shows md1 and md2 as UU, and SMART status is good.  Anyone have any suggestions?

Thx

Mike

[stephdl]

which smeserver version you have ?

Every sunday at 4.00 the raid is looking after errors and you can see it by 'watch cat /proc/mdstat' but normally you must not have an email of this.
It is normal.

[TerryF]

Had to replace a system board, and ever since mdstat fires every Sunday, but the mdstat  shows md1 and md2 as UU, and SMART status is good.  Anyone have any suggestions?

See info in wiki HERE

[mike_mattos]

smartctl for BOTH drives say no errors reported.   NO EMAILS about RAID errors.   Both partitions are UU in resync.

But it resyncs every week!

Version 8.1 and up to date.

'watch cat /proc/mdstat' tells me a rebuild is happening, and again NO EMAILS about a raid failure.  I'm stumped

[TerryF]

Did you read the wiki link?  "Periodic scrub of RAID arrays"   pretty much says it all

"Periodic scrub of RAID arrays

A Periodic scrub of RAID arrays (weekly raid check) is performed every week on Sunday at 04:22 local time, refer Bugzilla:3535 and Bugzilla:6160 for more information.

Theses operations are logged, however, no emails will be sent to admin as of the release of packages associated with Bug #6160 or the release of the 8.1 ISO."

[mike_mattos]

Does 'periodic scrub'  mean the array is rebuilt even if there are no problems?   

Curiously, the machine was rebooted during the rebuild, and the rebuild did NOT repeat, raid status just says all raid devices are in clean state.

So why did it resync on Sunday at 04:22 instead of saying all is well?

[stephdl]

Because mdadm is looking after bad clusters on your raid....if nothing is found then you win....if something is found, it is better to replace or repair your HD.

[mike_mattos]

what is confusing me is that the speed of the verify seems to be the same as the speed of the resync, suggesting the data is being written out to the drives. (compared to the speed of a read only compare)

Perhaps the actual resync output could have different wording than the weekly comparison when looking at cat /proc/mdstat .

[stephdl]

Well, we don't modify this rpm ('mdadm') it comes from centos6/EL6, we can not do something, only a bug report to redhat can change the wording.

Code: [Select]

[root@sme9 ~]# rpm -qf  /etc/cron.d/raid-check
mdadm-3.3-6.el6_6.1.i686
[root@sme9 ~]# rpm -qf /usr/sbin/raid-check
mdadm-3.3-6.el6_6.1.i686