Topic: RAID problem with 2 SCSII and 1 ATA

[barame]

Dear All,
I have the strange problem with my SME. First time I install SME 7.1 into 2 SCSII harddisk size each 18 GB.on Adaptec AIC7880 and one ATA 120 GB. SME Server see the whole space is 36 GB.
What RAID of this configuration. It automatic use RAID5 for me or not.
Problem is it working until my data growup to about 18 GB and then it stop working and can't start SME. Also when I try to re-install by upgrade from CD SME 7.2 it can't upgrade.
I wan't to access the data inside. How to get it from RAID.

Barame

[Reinhold]

Short explanation: SME can ONLY USE SAME SIZE DISKS !!!

Long version:
...with larger disks SME  will always take only the smallest size as common denominator...
...in your case it will accept your ATA120Gb as 18Gb only

SME then generates two Raid devices md1 and md2
(with a grain of salt):
- md1 is for system  i.e. boot and a Raid1
- md2 contains your data:  
 
With your 3 disks SME will generate a Raid5 array: 18+18+18GB
In a Raid5 usable size is (n-1) or in your case  2*18=36Gb

"Only 18Gb usable space" seems strange: please give result of: df -h (or install disk usage contrib)

"The data inside" - when you mean to say "the data in ibays" is located in "/home/e-smith/files/ibays" you can copy it to & from any network share

Barame ... i think you have to read the manual...
(your english is good enough and you are on the verge of loosing data)

Regards
Reinhold

P.S.:  ...nothing in your case is related to SCSI vs ATA...btw...

[slords]

With your 3 disks SME will generate a Raid5 array: 18+18+18GB
In a Raid5 usable size is (n-1) or in your case  2*18=36Gb

This isn't quite correct.  In SME 7.0 it would generate as stated above.  In 7.1  there was a bug and it would install in raid-1.  In 7.2 it will install as raid-1 with one hot-spare.

[barame]

Dear Reinhold
Thank you for your explanation about RAID in SME. What I want to do is I want to recover email in SME that I use IMAP for all users. and now This SME can't be us. I setup new SME in new harddisk and working with 1 hdd. Is it possible to plugin SCSII(one or two of them) and mount hdd in new SME as another mount location. Then I can move data uot from old scsii.

BR,
Barame

[TrevorB]

Is it possible to plugin SCSII(one or two of them) and mount hdd in new SME as another mount location. Then I can move data uot from old scsii.

Yes, just:

create a directory (eg. mkdir /tmp/olddisk)
mount your drive (eg. mount /dev/sda1 /tmp/olddisk)
copy whatever you want across
after that (when you have verified that all the data is across), you can add these drives into the smeserver (

AddExtraHardDisk)[/list]This assumes that the disks are formatted with a filesystem that smeserver understands.

Trevor B

[Reinhold]

In 7.1  there was a bug and it would install in raid-1.

...which does explain Barame's finding "of 18Gb only" quite nicely   (and Shad should know - he most likely coded it)

imap data recovery:
(1) Trevor gave you a good how to - methinks
(2nd proposal) In cases where the OS itself is down and I want the data from a single HD on another box I usually just put that HD into an external USB case ...
(actually I use an "interface only USB plug") - works and turnaround is faster considering that no setup (no "screwing" is needed

Regards
Reinhold

[barame]

Dear All
Thank you for all help. Now I can mount old SCSII disk but I still can't see my data from user profile. I can see only the boot partition and detail as below.

[root@web olddisk2]# ls
config-2.6.9-34.EL              lost+found
config-2.6.9-34.ELsmp           System.map-2.6.9-34.EL
config-2.6.9-42.0.10.EL         System.map-2.6.9-34.ELsmp
config-2.6.9-42.0.10.ELsmp      System.map-2.6.9-42.0.10.EL
grub                            System.map-2.6.9-42.0.10.ELsmp
initrd-2.6.9-34.EL.img          vmlinuz-2.6.9-34.EL
initrd-2.6.9-34.ELsmp.img       vmlinuz-2.6.9-34.ELsmp
initrd-2.6.9-42.0.10.EL.img     vmlinuz-2.6.9-42.0.10.EL
initrd-2.6.9-42.0.10.ELsmp.img  vmlinuz-2.6.9-42.0.10.ELsmp
[root@web olddisk2]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/md3              110G  1.6G  103G   2% /
/dev/md1               99M  8.6M   85M  10% /boot
none                  570M     0  570M   0% /dev/shm
/dev/sda1              99M  9.4M   85M  11% /tmp/olddisk
/dev/sdb1              99M   19M   75M  21% /tmp/olddisk2
[root@web olddisk2]#

I did somthing wrong or where is the mount point instead of /dev/sda1 or sdb1

Best regards,
Barame

[TrevorB]

[root@web olddisk2]# ls
config-2.6.9-34.EL              lost+found
config-2.6.9-34.ELsmp           System.map-2.6.9-34.EL
config-2.6.9-42.0.10.EL         System.map-2.6.9-34.ELsmp
config-2.6.9-42.0.10.ELsmp      System.map-2.6.9-42.0.10.EL
grub                            System.map-2.6.9-42.0.10.ELsmp
initrd-2.6.9-34.EL.img          vmlinuz-2.6.9-34.EL
initrd-2.6.9-34.ELsmp.img       vmlinuz-2.6.9-34.ELsmp
initrd-2.6.9-42.0.10.EL.img     vmlinuz-2.6.9-42.0.10.EL
initrd-2.6.9-42.0.10.ELsmp.img  vmlinuz-2.6.9-42.0.10.ELsmp

This is all the new /boot directory for smeserver - so it has definitely been overwritten.

[root@web olddisk2]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/md3              110G  1.6G  103G   2% /
/dev/md1               99M  8.6M   85M  10% /boot

OK, can you please do a 'mdadm --detail /dev/md3'

/dev/sda1              99M  9.4M   85M  11% /tmp/olddisk
/dev/sdb1              99M   19M   75M  21% /tmp/olddisk2

So what does a ls /tmp/olddisk give you?

I did somthing wrong or where is the mount point instead of /dev/sda1 or sdb1

With the multiple tries at trying to install etc., you may have scrubbed both disks.

What is the output of 'fdisk -l /dev/sda' and 'fdisk -l /dev/sdb'?

Try and answer all of the above questions and we'll see.

Trevor B

[barame]

Dear Trevorb
The all story is after I have 2x18GB SCSII and IDE 120GB running on Raid5 with total space 36GB. Then the system crash can't boot up.
1.I disconnect SCSI ID1 and IDE from the server and setup SME as upgrade in SCSI ID0 but the upgrade process can't be completed.
2.I setup as new SME in SCSI ID0 but the SCSI can't boot.
3.I take all disk out from this server. Put the new IDE 120 GB. and setup as new SME 7.2 and restore data from backup. It working fine. but some data loss may be email from last month as data in backup I made it from last time.
4. I put in SCSI ID0 and ID1 into this system and run mount command to /dev/sda1 and /dev/sdb1

From what I did all Data in SCSI ID0 was totally lost. but the SCSI ID1 and old IDE120 GB. should be inside.

From your question

[root@web ~]# mdadm --detail /dev/md3
/dev/md3:
        Version : 00.90.01
  Creation Time : Wed Aug  1 18:38:20 2007
     Raid Level : raid1
     Array Size : 116591616 (111.19 GiB 119.39 GB)
    Device Size : 116591616 (111.19 GiB 119.39 GB)
   Raid Devices : 2
  Total Devices : 1
Preferred Minor : 3
    Persistence : Superblock is persistent

    Update Time : Wed Aug  8 12:51:41 2007
          State : clean, degraded
 Active Devices : 1
Working Devices : 1
 Failed Devices : 0
  Spare Devices : 0

           UUID : 4cb869ea:a12c6ccd:cfea1849:c6b7e53c
         Events : 0.68831

    Number   Major   Minor   RaidDevice State
       0       3        2        0      active sync   /dev/hda2
       1       0        0        -      removed
[root@web ~]#

This is new IDE 120 GB harddisk.

from ls /tmp/olddisk this is all new setup in SCSI ID0

[root@web ~]# ls /tmp/olddisk
config-2.6.9-55.0.2.EL  initrd-2.6.9-55.0.2.EL.img  System.map-2.6.9-55.0.2.EL
grub                    lost+found                  vmlinuz-2.6.9-55.0.2.EL
[root@web ~]# fdisk -l /dev/sda

Disk /dev/sda: 18.2 GB, 18200739840 bytes
255 heads, 63 sectors/track, 2212 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1          13      104391   fd  Linux raid autodetect
/dev/sda2              14        2212    17663467+  fd  Linux raid autodetect
[root@web ~]# fdisk -l /dev/sdb

Disk /dev/sdb: 18.2 GB, 18200739840 bytes
255 heads, 63 sectors/track, 2212 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *           1          13      104391   fd  Linux raid autodetect
/dev/sdb2              14        2212    17663467+  fd  Linux raid autodetect
[root@web ~]#

Need Urgent help please

Barame

[barame]

Urgent need please
how to mount data in /dev/sdb2 I can mount by using
mount /dev/sdb1 /tmp/olddisk
But When I use command
mount /dev/sdb2 /tmp/olddisk2
It said
mount: you must specify the filesystem type
I think my home data is in /dev/sdb2 partition

Barame

[TrevorB]

mount /dev/sdb2 /tmp/olddisk2
It said
mount: you must specify the filesystem type
I think my home data is in /dev/sdb2 partition

Then you look like you have a problem....

If it can't autodetect the filesystem type, then it is probably not formatted. You can try:

Code: [Select]

mount /dev/sdb2 /tmp/olddisk2 -t ext3but if that doesn't work, I would suggest that this partition has been wiped.

In that case, you're well beyond my understanding of how to recover your data.

Trevor B

[Reinhold]

Barame,

Up there we concluded that you had a Raid1 installed
...and that you are using a single disk from that Raid1

If you have done more (that md3 label up there is still puzzeling me)
and the current disk is in reality "just1 disk" of a "3disk - 1missing"-degraded-Raid5
you do need to reform a (degraded) raid with the other disk...
If you do not have that disk "functional" any more I am afraid you are "out of options"
There is no way you can assemble your Raid5 data with 2 disks missing...

Regards
Reinhold

...x-ing fingers the above is wrong and adding ext3 did it.
 

[Gaston94]

Hi,
you have a SME 7 installtion which was performed with LVM and RAID. You can access sdb1 because this part is not included in the LVM stuff

Urgent need please
how to mount data in /dev/sdb2 I can mount by using
mount /dev/sdb1 /tmp/olddisk
But When I use command
mount /dev/sdb2 /tmp/olddisk2
It said
mount: you must specify the filesystem type

But to access the data on sdb2, you have to restart the old raid array and activate the LVM structure.
I remember there was some Faq/howto but cannot retrieve it yet  But most important part for the story is there

G.