Software RAID1 problems

Software RAID1 problems

[Jean-Rene Cormier]

Well since I had some trouble with raidtools giving me a Segfault when I
tried to create an array, I decided to try mdadm, I can create an array
and format it and mount it afterwards but after I unmount it I can't
mount it again. I get an error saying unknown FS type or something like
that. When I do an mdadm --detail /dev/md0 in the state it says dirty,
no-error, shouldn't that be clean, no-error? How can I get that array to
that state? I tried recreating that array like 5 times already and still
the same thing. btw I'm running Linux 2.4.20 with glibc-2.3.1 if that
changes anything.

Jean-Rene Cormier

Re: Software RAID1 problems

[Neil Brown]

On  February 20, jean-rene.cormier@cipanb.ca wrote:
> Well since I had some trouble with raidtools giving me a Segfault when I
> tried to create an array, I decided to try mdadm, I can create an array
> and format it and mount it afterwards but after I unmount it I can't
> mount it again. I get an error saying unknown FS type or something like
> that. When I do an mdadm --detail /dev/md0 in the state it says dirty,
> no-error, shouldn't that be clean, no-error? How can I get that array to
> that state? I tried recreating that array like 5 times already and still
> the same thing. btw I'm running Linux 2.4.20 with glibc-2.3.1 if that
> changes anything.

It is always dirty when assembled and writable.  This is normal.

If you switch to read-only, it should should clean.
If you stop the array and use --examine on the devices, they should
show clean.

I cannot explain the mounting problem.  Maybe if you give more
specifics....

NeilBrown

Re: Software RAID1 problems

[Jean-Rene Cormier]

----- Original Message -----
From: "Neil Brown" <neilb@cse.unsw.edu.au>
To: "Jean-Rene Cormier" <jean-rene.cormier@cipanb.ca>
Cc: <linux-raid@vger.kernel.org>
Sent: Thursday, February 20, 2003 8:11 PM
Subject: Re: Software RAID1 problems


> On  February 20, jean-rene.cormier@cipanb.ca wrote:
> > Well since I had some trouble with raidtools giving me a Segfault when I
> > tried to create an array, I decided to try mdadm, I can create an array
> > and format it and mount it afterwards but after I unmount it I can't
> > mount it again. I get an error saying unknown FS type or something like
> > that. When I do an mdadm --detail /dev/md0 in the state it says dirty,
> > no-error, shouldn't that be clean, no-error? How can I get that array to
> > that state? I tried recreating that array like 5 times already and still
> > the same thing. btw I'm running Linux 2.4.20 with glibc-2.3.1 if that
> > changes anything.
>
> It is always dirty when assembled and writable.  This is normal.
>
> If you switch to read-only, it should should clean.
> If you stop the array and use --examine on the devices, they should
> show clean.
>
> I cannot explain the mounting problem.  Maybe if you give more
> specifics....

Well what I did was:

mdadm -Cv -l2 -n2 /dev/hde1 /dev/hdg1 /dev/md0
mke2fs -j /dev/md0
mount /dev/md0 /home

then I unmounted it and when I tried to mount it again right after
unmounting it I got the Wrong FS type error and I get a bunch of error
message on the console.

Jean-Rene Cormier

Re: Software RAID1 problems

[Neil Brown]

On Thursday February 20, jean-rene.cormier@cipanb.ca wrote:
> >
> > I cannot explain the mounting problem.  Maybe if you give more
> > specifics....
> 
> Well what I did was:
> 
> mdadm -Cv -l2 -n2 /dev/hde1 /dev/hdg1 /dev/md0
> mke2fs -j /dev/md0
> mount /dev/md0 /home
> 
> then I unmounted it and when I tried to mount it again right after
> unmounting it I got the Wrong FS type error and I get a bunch of error
> message on the console.

If that is really what you did, no wonder it didn't work.  The md
device must be listed first, not last.

However if it is a typo, and you really listed /dev/md0 first, it is
strange.  Maybe if you could include the 'bunch of error messages' - 
   dmesg > /tmp/afile
should get them for you.

NeilBrown

Re: Software RAID1 problems

[Jean-Rene Cormier]

----- Original Message -----
From: "Neil Brown" <neilb@cse.unsw.edu.au>
To: "Jean-Rene Cormier" <jean-rene.cormier@cipanb.ca>
Cc: <linux-raid@vger.kernel.org>
Sent: Sunday, February 23, 2003 6:40 PM
Subject: Re: Software RAID1 problems


> On Thursday February 20, jean-rene.cormier@cipanb.ca wrote:
> > >
> > > I cannot explain the mounting problem.  Maybe if you give more
> > > specifics....
> >
> > Well what I did was:
> >
> > mdadm -Cv -l2 -n2 /dev/hde1 /dev/hdg1 /dev/md0
> > mke2fs -j /dev/md0
> > mount /dev/md0 /home
> >
> > then I unmounted it and when I tried to mount it again right after
> > unmounting it I got the Wrong FS type error and I get a bunch of error
> > message on the console.
>
> If that is really what you did, no wonder it didn't work.  The md
> device must be listed first, not last.
>
> However if it is a typo, and you really listed /dev/md0 first, it is
> strange.  Maybe if you could include the 'bunch of error messages' -
>    dmesg > /tmp/afile
> should get them for you.

Yes that was a typo sorry. I did reformat /dev/md0 and a when it started
formating it gave me those errors again. I did a little searching and the
problem is not in the raid array, but I'm still not sure what is causing the
problem. The error was:

hde: status error: status=0x58 { DriveReady SeekComplete DataRequest }
hde: drive not ready for command
hde: status timeout: status=0xf0 { Busy }
hde: drive not ready for command
ide2: reset: success

I get a bunch of those. I removed the drive and I reformated the array and
it works fine but I guess that if I put that drive back in it'll start doing
it again. Could it be a bad drive or something? If it is what's the use of
having a RAID-1 array if it corrupts your file system when a drive is bad?

Jean-Rene Cormier

Re: Software RAID1 problems

[Alvin Oga]

On Sat, 1 Mar 2003, Jean-Rene Cormier wrote:

> Yes that was a typo sorry. I did reformat /dev/md0 and a when it started
> formating it gave me those errors again. I did a little searching and the
> problem is not in the raid array, but I'm still not sure what is causing the
> problem. The error was:
> 
> hde: status error: status=0x58 { DriveReady SeekComplete DataRequest }
> hde: drive not ready for command
> hde: status timeout: status=0xf0 { Busy }
> hde: drive not ready for command
> ide2: reset: success

make sure you have the same speed disks on the same ide cable
	- best to use only one ide drive per cable
	- do NOT mix ata-33 w/ ata-66 w/ ata-100 w/ ata-133 on the same
	cable

use better and SHORTER ide cables ??   ( should be 80 conductor )

make sure you do NOT have kinks in the ide cable ( keep it straight )

make sure your kernel has the ide chipset support enabled for your
motherboard


-- otherwise... you have a failing or dead disk drive...
	-- if its an ibm deskstar... get rid of it...

c ya
alvin

Re: Software RAID1 problems

[Jean-Rene Cormier]

----- Original Message -----
From: "Alvin Oga" <aoga@Maggie.Linux-Consulting.com>
To: "Jean-Rene Cormier" <jean-rene.cormier@cipanb.ca>
Cc: <linux-raid@vger.kernel.org>
Sent: Saturday, March 01, 2003 9:08 PM
Subject: Re: Software RAID1 problems


>
> On Sat, 1 Mar 2003, Jean-Rene Cormier wrote:
>
> > Yes that was a typo sorry. I did reformat /dev/md0 and a when it started
> > formating it gave me those errors again. I did a little searching and
the
> > problem is not in the raid array, but I'm still not sure what is causing
the
> > problem. The error was:
> >
> > hde: status error: status=0x58 { DriveReady SeekComplete DataRequest }
> > hde: drive not ready for command
> > hde: status timeout: status=0xf0 { Busy }
> > hde: drive not ready for command
> > ide2: reset: success
>
> make sure you have the same speed disks on the same ide cable
> - best to use only one ide drive per cable
> - do NOT mix ata-33 w/ ata-66 w/ ata-100 w/ ata-133 on the same
> cable

Both drives are new 60Gb Maxtor DiamondMax Plus 9 on separate IDE channels
on a Promise Ultra TX2 IDE card

> use better and SHORTER ide cables ??   ( should be 80 conductor )

I did switch the cable on the drives to see if it would make a difference
and it's still the same drive who gave me those errors so I guess it's not
the cables.

> make sure you do NOT have kinks in the ide cable ( keep it straight )
>
> make sure your kernel has the ide chipset support enabled for your
> motherboard

I noticed yesterday that my kernel didn't have IDE support for the Promise
Ultra compiled in so I recompiled the kernel with it. I'll put the drive
back in the computer and try it again to see if it works better

> -- otherwise... you have a failing or dead disk drive...
> -- if its an ibm deskstar... get rid of it...

I'll try this drive in another computer to see what it'll say.

Thanks

Jean-Rene Cormier

ANNOUNCE: mdadm 2.1 - A tool for managing Soft RAID under Linux

[Neil Brown]

I am pleased to announce the availability of
   mdadm version 2.1

It is available at the usual places:
   http://www.cse.unsw.edu.au/~neilb/source/mdadm/
and (soon)
   http://www.{countrycode}.kernel.org/pub/linux/utils/raid/mdadm/

mdadm is a tool for creating, managing and monitoring
device arrays using the "md" driver in Linux, also
known as Software RAID arrays.

Release 2.1 fixes a few problems found after the release of 2.0.
If you are using raid10 or version-1 superblocks and upgrade is
recommended.

Specifically:
    -   Fix assembling of raid10 array when devices are missing.
	  mdadm now correctly detects if a array is workable or not
	  depending on which devices are present, and so will correctly
	  handle "--assemble --force"  if multiple devices have failed.
    -   Report raid10 layout in --examine output.
    -   Fix assembling of arrays that use the version-1 superblock and
	  have spares.  Previously the spares would be ignored.
    -   Fix bug so that multiple drives can be re-added at once.
    -   Fix problem with hot-adding a bitmap to version-1-superblock
	  arrays.


Development of mdadm is sponsored by
 SUSE Labs, Novell Inc.

NeilBrown  12th September 2005

Software RAID1 problems

[Andy Dawson]

I've had some problems recently with a RAID1 setup on a couple of Seagate 
SCSI disks.  I've setup the following partition scheme:

/dev/md0 swap (/dev/sda1 and /dev/sdb1)
/dev/md1 /    (/dev/sda2 and /dev/sdb2)

md1 functions fine, and has so far not given me any problems at all.  Md0 
however keeps dropping /dev/sdb1 from the array.  As I've set the system up 
to let me know via e-mail should anything occur, I do receive notification 
when sdb1 is kicked from the array.

Hot adding sdb1 back into the array works fine, until the next 
'degradedarray event'.

I've tested the disks with the Seagate tools (no problems), badblocks in 
read-only mode and in write mode as well (recreating md0 afterwards), all 
with no problems reported.

Has anyone any idea what might be causing the degradedarray event to 
occur?

Andy

Dr. Andy Dawson
A.J.Dawson@Bradford.ac.uk
http://www.mossie.org
http://www.museum-explorer.org.uk

  Never attribute to malice that which can adequately be explained by stupidity

Re: Software RAID1 problems

[Neil Brown]

On Monday September 12, A.J.Dawson@Bradford.ac.uk wrote:
> 
> Has anyone any idea what might be causing the degradedarray event to 
> occur?

What kernel messages do you get when it happens (in /var/log/kern.log
on wherever your distro puts them)?

NeilBrown

Re: Software RAID1 problems

[Andy Dawson]

After a fairly long time without seeing the following message:

"This is an automatically generated mail message from mdadm
running on servername

A DegradedArray event had been detected on md device /dev/md0.

Faithfully yours, etc."

It reared its head again today.

I've checked dmesg and /var/log/messages and the only thing I've found so 
far is the following in dmesg:

md: considering sda1 ...
md:  adding sda1 ...
md: created md0
md: bind<sda1>
md: running: <sda1>
raid1: raid set md0 active with 1 out of 2 mirrors

md1 works fine and md sets md1 active with 2 out of 2 mirrors.

/dev/sdb2 is the device that appears to be causing the problem, so can 
anyone tell me what might be happening?  Any ideas why it might not be 
included in the array at boot time? (I've tried Google but found no 
answers yet).

Any help would be appreciated!

Regards
Andy



On Tue, 13 Sep 2005, Neil Brown wrote:

> On Monday September 12, A.J.Dawson@Bradford.ac.uk wrote:
>>
>> Has anyone any idea what might be causing the degradedarray event to
>> occur?
>
> What kernel messages do you get when it happens (in /var/log/kern.log
> on wherever your distro puts them)?
>
> NeilBrown
>

Dr. Andy Dawson
A.J.Dawson@Bradford.ac.uk
http://www.mossie.org
http://www.museum-explorer.org.uk

  Never attribute to malice that which can adequately be explained by stupidity

Re: Software RAID1 problems

[Neil Brown]

On Friday September 23, A.J.Dawson@Bradford.ac.uk wrote:
> After a fairly long time without seeing the following message:
> 
> "This is an automatically generated mail message from mdadm
> running on servername
> 
> A DegradedArray event had been detected on md device /dev/md0.
> 
> Faithfully yours, etc."
> 
> It reared its head again today.
> 
> I've checked dmesg and /var/log/messages and the only thing I've found so 
> far is the following in dmesg:
> 
> md: considering sda1 ...
> md:  adding sda1 ...
> md: created md0
> md: bind<sda1>
> md: running: <sda1>
> raid1: raid set md0 active with 1 out of 2 mirrors
> 
> md1 works fine and md sets md1 active with 2 out of 2 mirrors.
> 
> /dev/sdb2 is the device that appears to be causing the problem, so can 
> anyone tell me what might be happening?  Any ideas why it might not be 
> included in the array at boot time? (I've tried Google but found no 
> answers yet).
> 
> Any help would be appreciated!

Looks a lot like sdb2 (sure it isn't sdb1?) isn't type 'FD'.
NeilBrown