[BlueOnyx:03017] RAID resynch every week

[Jeff Folk]

Bump... Does anyone have any ideas about this? The weekly raid-check  
routine continues to rebuild my array (md0 and md2) every time it is  
run... Did it this morning with the weekly cron, then I performed a  
mdchk immediately before running the 99-raid-check job manually just  
now:

> [root at box1 ~]# /etc/init.d/mdchk
>
> -- Reading system configuration from /etc/raidtab
>    /dev/md0 [raid1]: <sda1> <sdd1> <sdc1> <sdb1>
>    /dev/md1 [raid1]: <sdc2> <sda2>
>    /dev/md2 [raid1]: <sdd2> <sdb2>
>
> -- Reading current status from /proc/mdstat
>    /dev/md0 [raid1]: <sda1> <sdd1> <sdc1> <sdb1>  (ACTIVE)
>    /dev/md1 [raid1]: <sdc2> <sda2>  (ACTIVE)
>    /dev/md2 [raid1]: <sdd2> <sdb2>  (ACTIVE)
> -- RAID status: OK

Now it is syncing AGAIN... Even though the RAID status is OK?

> Original message:
>
> I'm getting concerned... Should I be?
>
> cron.weekly apparently runs a 99-raid-check job every Sunday morning  
> at 4:22 AM, and my RAID0+1 array gets md0 and md2 rebuilt every week  
> (according to the Active Monitor e-mails)... /var/log/messages looks  
> like this every time:
>
> Nov 22 04:22:01 box1 kernel: md: syncing RAID array md0
> Nov 22 04:22:01 box1 kernel: md: minimum _guaranteed_ reconstruction  
> speed: 35000 KB/sec/disc.
> Nov 22 04:22:01 box1 kernel: md: using maximum available idle IO  
> bandwidth (but not more than 200000 KB/sec) for reconstruction.
> Nov 22 04:22:01 box1 kernel: md: using 128k window, over a total of  
> 256896 blocks.
> Nov 22 04:22:01 box1 kernel: md: delaying resync of md2 until md0  
> has finished resync (they share one or more physical units)
> Nov 22 04:22:07 box1 kernel: md: md0: sync done.
> Nov 22 04:22:07 box1 kernel: md: syncing RAID array md2
> Nov 22 04:22:07 box1 kernel: md: minimum _guaranteed_ reconstruction  
> speed: 35000 KB/sec/disc.
> Nov 22 04:22:07 box1 kernel: md: using maximum available idle IO  
> bandwidth (but not more than 200000 KB/sec) for reconstruction.
> Nov 22 04:22:07 box1 kernel: md: using 128k window, over a total of  
> 312311552 blocks.
> Nov 22 04:22:07 box1 kernel: RAID1 conf printout:
> Nov 22 04:22:07 box1 kernel:  --- wd:4 rd:4
> Nov 22 04:22:07 box1 kernel:  disk 0, wo:0, o:1, dev:sda1
> Nov 22 04:22:07 box1 kernel:  disk 1, wo:0, o:1, dev:sdb1
> Nov 22 04:22:07 box1 kernel:  disk 2, wo:0, o:1, dev:sdc1
> Nov 22 04:22:07 box1 kernel:  disk 3, wo:0, o:1, dev:sdd1
>
> Resynch completes every time, and I never get any other  
> notifications of degraded disk integrity. Are there any other  
> historical logs I can explore to see what is going on? What should I  
> be looking for??
>
> Many thanks in advance.
>
> Regards;
> Jeff Folk