RAID 1 question: drive just dropped out...

[Don Levey-7]

When powering up my machine yesterday, I found that I have a degraded
RAID-I array.  Without posting all of the logs here, the operant lines
seem to be:

        kicking non-fresh sdb1 from array!
        ...
        unbind<sdb1>

smartctl says that /dev/sdb is "healthy".  Is this just an instance
where I should try to manually ass the sdb partitions with mdadm, or is
there something else I should try first?

  -Don

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
Discuss mailing list
Discuss@...
http://lists.blu.org/mailman/listinfo/discuss

[Don Levey-7]

Don Levey wrote:

> When powering up my machine yesterday, I found that I have a degraded
> RAID-I array.  Without posting all of the logs here, the operant lines
> seem to be:
>
>     kicking non-fresh sdb1 from array!
>     ...
>     unbind<sdb1>
>
> smartctl says that /dev/sdb is "healthy".  Is this just an instance
> where I should try to manually ass the sdb partitions with mdadm, or is
> there something else I should try first?
>
>  -Don
>

I'm adding them now - they say they're synching up.  But what would
cause this?  Should I be worried?

  -Don

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
Discuss mailing list
Discuss@...
http://lists.blu.org/mailman/listinfo/discuss

[Dan Ritter-2]

On Thu, Jul 03, 2008 at 11:15:59AM -0400, Don Levey wrote:

> Don Levey wrote:
> >When powering up my machine yesterday, I found that I have a degraded
> >RAID-I array.  Without posting all of the logs here, the operant lines
> >seem to be:
> >
> >    kicking non-fresh sdb1 from array!
> >    ...
> >    unbind<sdb1>
> >
> >smartctl says that /dev/sdb is "healthy".  Is this just an instance
> >where I should try to manually ass the sdb partitions with mdadm, or is
> >there something else I should try first?
> >
> > -Don
> >
>
> I'm adding them now - they say they're synching up.  But what would
> cause this?  Should I be worried?

I've had this happen when we lost power unexpectedly. After the
array is mounted and synched, no further problems.

-dsr-

--
http://tao.merseine.nu/~dsr/eula.html is hereby incorporated by reference.

When freedom gets lots of exercise, it protects itself.

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
Discuss mailing list
Discuss@...
http://lists.blu.org/mailman/listinfo/discuss

[Matthew Gillen]

Don Levey wrote:

> Don Levey wrote:
>> When powering up my machine yesterday, I found that I have a degraded
>> RAID-I array.  Without posting all of the logs here, the operant lines
>> seem to be:
>>
>>     kicking non-fresh sdb1 from array!
>>     ...
>>     unbind<sdb1>
>>
>> smartctl says that /dev/sdb is "healthy".  Is this just an instance
>> where I should try to manually ass the sdb partitions with mdadm, or
>> is there something else I should try first?
>>
>>  -Don
>>
>
> I'm adding them now - they say they're synching up.  But what would
> cause this?  Should I be worried?

I had that happen once.  I just happened to look at /proc/mdstat and noticed
that one of my drives was missing from all the RAID arrays it was supposed to
be a part of.  No idea how long it had been like that.

I have smartd set up to send mail when something exciting happens with one of
my drives, but whatever happened wasn't bad enough for smartd to get concerned.

So now I periodically check /proc/mdstat just to be sure (but it's never
happened again).

Matt

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
Discuss mailing list
Discuss@...
http://lists.blu.org/mailman/listinfo/discuss

[Don Levey-7]

Matthew Gillen wrote:

> I have smartd set up to send mail when something exciting happens with
> one of my drives, but whatever happened wasn't bad enough for smartd to
> get concerned.
>

That's what alerted me - the machine came up sans /dev/sdb, so I got
messages for each partition.

Dan Ritter wrote:
 >
 > I've had this happen when we lost power unexpectedly. After the
 > array is mounted and synched, no further problems.
 >

That appears to have been the trigger here too.  When getting the old
server off of the table, I managed to snag the power cord for the new one.

The array is synching right now, no problems reported.  I love it when
something "just works," but I'm sure not used to it.

  -Don

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
Discuss mailing list
Discuss@...
http://lists.blu.org/mailman/listinfo/discuss

[Rich Braun]

Matthew Gillen <me@...> wrote:
> I have smartd set up to send mail when something exciting happens with one of
> my drives, but whatever happened wasn't bad enough for smartd to get
> concerned.

The command you want is:

  /sbin/mdadm --monitor --delay=300 --scan --daemonize /dev/md0

The file you want to create is /etc/mdadm.conf, mine is:

DEVICE    /dev/hda2 /dev/hdc2
ARRAY     /dev/md0 level=raid1 num-devices=2 UUID=c2cc6135:4e5560c7:6204121b:f92
a7259
          devices=/dev/hda2,/dev/hdc2
MAILADDR  admin

--
To generate the ARRAY line, invoke:

  mdadm --detail --scan

This information is in the man page for mdadm.

-rich

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
Discuss mailing list
Discuss@...
http://lists.blu.org/mailman/listinfo/discuss