Koozali.org: home of the SME Server

degraded raid array

Offline kdn954

  • 2
  • +0/-0
degraded raid array
« on: September 13, 2009, 10:14:26 PM »
I am receiving an e-mail stating that both of my drives on raid 1 array are degraded. How do I repair if both are degraded. Thanks

Offline byte

  • *
  • 2,183
  • +2/-0
Re: degraded raid array
« Reply #1 on: September 13, 2009, 10:38:47 PM »
I am receiving an e-mail stating that both of my drives on raid 1 array are degraded. How do I repair if both are degraded. Thanks

Have a read regarding the S.M.A.R.T tools to diagnose your drive and find out if you have a faulty drive, also use the manufacture harddrive tools to see if you have any bad sectors so on.

Once your confident the drive is safe you can resync using the Manage raid option in the console, their are also other ways too.
--[byte]--

Have you filled in a Bug Report over @ http://bugs.contribs.org ? Please don't wait to be told this way you help us to help you/others - Thanks!

Offline Stefano

  • *
  • 10,839
  • +2/-0
Re: degraded raid array
« Reply #2 on: September 13, 2009, 10:42:34 PM »
I am receiving an e-mail stating that both of my drives on raid 1 array are degraded. How do I repair if both are degraded. Thanks

Hi, welcome here

first of all: if you receive that mail it means that your server is running.. if both of your drivers were faulty, you would never have received it :wink:

Please take the time to read the documentation and to search the forum before posting: your question has ben posted many times

thank you

Stefano

Offline aussierob

  • **
  • 55
  • +0/-0
Re: degraded raid array
« Reply #3 on: September 30, 2009, 04:39:23 AM »
Please take the time to read the documentation and to search the forum before posting: your question has ben posted many times
thank you
Stefano

Hi Stefano,
I too am having small trouble with RAID, even after reading forum posts.
If you have comments on the below, I would welcome them: if not still cheers to you.
I installed SME with both drives installed (as the main doc I was reading didn't say to only do one):
that took a long time, but all seems well.
Both drives show in mdstat
Code: [Select]
md2 : active raid1 hda2[0] hdc2[1]
         976655488 blocks [2/2] [UU]
         [>.....] resync 0.2% (2590000/976655488) finish=13390min speed=1350k/sec
md1 : active raid1 hda1[0] hdc1[1]
         104320 blocks [2/2] [UU]
unused devices: <none>
It is a pair of 1TB Samsung drives and I found advise to slow them to 1.5Gb rather than 3Gb, but a couple of things still concern me...
1. 1350k.sec seems slow, but it is a background task, although I cannot figure how to speed it up. (it will take 200hrs! - but then I suppose it IS 1000,000Mb eh)
2. After a power failure longer than my UPS could handle, upon reboot the darned sync started again (it was 85% done): is there anyway to note a point of the build)
3. Maybe I should have had the big filesystem split into 2 (say 20GB and 900GB, as the larger will just be a common file storeage): but I would need ibays on the big one, which would not have /home..../files/ibays directory.
3. I cannot see a way to do anything else other than wait (I DID check the console after reading this post, but it only 'manual intervention required')
4. I did look up man mdadm, but that does not look for the fainthearted!

Kind Regards
Robert

Offline Stefano

  • *
  • 10,839
  • +2/-0
Re: degraded raid array
« Reply #4 on: September 30, 2009, 09:03:03 AM »
It is a pair of 1TB Samsung drives and I found advise to slow them to 1.5Gb rather than 3Gb, but a couple of things still concern me...
1. 1350k.sec seems slow, but it is a background task, although I cannot figure how to speed it up. (it will take 200hrs! - but then I suppose it IS 1000,000Mb eh)

yes.. it's too slow.. I suspect an hw issue.. your chipset could be not properly recognized.. colud you tell us something about your hw?

Quote
2. After a power failure longer than my UPS could handle, upon reboot the darned sync started again (it was 85% done): is there anyway to note a point of the build)

AFAIK there's a way to achieve it, but it needs a kernel patch and so it isn't available on SME
question: did you configure NUT for your ups?

Quote
3. Maybe I should have had the big filesystem split into 2 (say 20GB and 900GB, as the larger will just be a common file storeage): but I would need ibays on the big one, which would not have /home..../files/ibays directory.
3. I cannot see a way to do anything else other than wait (I DID check the console after reading this post, but it only 'manual intervention required')

you didn't do anything wrong :-)

ciao

Offline byte

  • *
  • 2,183
  • +2/-0
Re: degraded raid array
« Reply #5 on: September 30, 2009, 11:31:21 AM »
Both drives show in mdstat
Code: [Select]
md2 : active raid1 hda2[0] hdc2[1]
         976655488 blocks [2/2] [UU]
         [>.....] resync 0.2% (2590000/976655488) finish=13390min speed=1350k/sec
md1 : active raid1 hda1[0] hdc1[1]
         104320 blocks [2/2] [UU]
unused devices: <none>
It is a pair of 1TB Samsung drives and I found advise to slow them to 1.5Gb rather than 3Gb, but a couple of things still concern me...

Are you sure you can tweak 1.5Gb from 3Gb as the mdstat you show use say your using IDE drives not SATA drives ?!

Quote
1. 1350k.sec seems slow, but it is a background task, although I cannot figure how to speed it up. (it will take 200hrs! - but then I suppose it IS 1000,000Mb eh)

If your mdstat is correct and your using IDE drives then thats about the speed I'd expect.
--[byte]--

Have you filled in a Bug Report over @ http://bugs.contribs.org ? Please don't wait to be told this way you help us to help you/others - Thanks!

Offline aussierob

  • **
  • 55
  • +0/-0
Re: degraded raid array
« Reply #6 on: October 01, 2009, 03:37:30 AM »
yes.. it's too slow.. I suspect an hw issue.. your chipset could be not properly recognized.. colud you tell us something about your hw?
ciao
G"day
Motherboard is GIGABYTE GA-G31M-ES2L
E6300 dual 2.6GHz, 2GB memory
2 x Samsung HD103UJ 1TB disks for RAID

A look through /var/log/dmesg shows (well done Stephano they are IDE!)
Code: [Select]
Uniform Milt-platform E-IDE driver Revision: 7.00alpha2
Probing IDE interface ide0...
hda: SAMSUNG HD103UJ, ATA DISK drive
hdb: TSSTcorp CDDVDW SH-S223B, ATAPI CD/DVD-ROM drive
Probing IDE interface ide1...
hdc: SAMSUNG HD103UJ, ATA DISK drive
and further down at the raid details...
Code: [Select]
md: raid1 personality registered as nr 3
md: md1 stopped.
md: bind<hdc1>
md: bind<hda1>
raid1: raid set to md1 active with 2 out of 2 mirrors
I also rechecked that when I use the downloaded Samsung program, booted from CD, to change the drive max-speed limit from 3Gb to1.5Gb it see drives by type on primary master and secondary master (which I did not register as 'bad' as this is my 1st SATA use).

The BIOS and CMOS settings shows IDE drives per above (BIOS is Award Modular BIOS v6.00PG, c1984-2009, Award Software Inc)

I suppose that other than the slow raid-sync I don't care too much as long as there are no on-going issues (which DOES concern me).
So it sounds like I need an improved BOIS or replacement motherboard?

Thanks for your thoughts on this Stephano
Regards Robert


Offline Stefano

  • *
  • 10,839
  • +2/-0
Re: degraded raid array
« Reply #7 on: October 01, 2009, 08:57:09 AM »
Hi

check (using the manual) your bios setup as you should see your hds as sata, not ide
hint: page 42.. set your sata controller in enhanced mode

hth

Offline aussierob

  • **
  • 55
  • +0/-0
Re: degraded raid array
« Reply #8 on: October 02, 2009, 01:19:33 AM »
check (using the manual) your bios setup as you should see your hds as sata, not ide
hint: page 42.. set your sata controller in enhanced mode
many MANY Thanks Stephano !
I found the parameter and now she goes like a rocket (my 1st SATA experience) !
mdstat shows 115MB/sec 147mins to go !
and everything has booted from the changed address (I will 'alter' the config though just to get it to rework it in case)
Kind Regards Robert