Re: [gentoo-server] Asking for advices after harddisk crash.

[Penghui Wang]

Thanks very much, David,

Could you tell me why you said that the problem is one of the drivers and
not the controller? In my opinion, i think the filesystem has been broken by
the force online and force check operations. So i am afraid if i remove one
of the disks, then the current stuation will be changed, and it would be
harder to recover the filesystem.

The extened partition sda4, contained two logic partitions. swap and /home.
Now both of them disappeared, and there is a wrong beginning and end
cylinker sda5 there.

If i create the two partition manully as the very same both begining and
end as before, is it helpfull. Or there are some other partition scaning
program i could use?

Thank very much again.

Penghui Wang




On 9/17/07, David Gardner <david@×××××××××.net> wrote:
>
> Assuming that the problem is one of the drives and not the controller,
> you could try booting with only two of the disks online. If the Raid
> controller is the problem you may need another controller of the same
> model to re-construct the array.
>
> Penghui Wang wrote:
> > Hello, list,
> >
> > Recently, one of our servers has been crashed because of some harddisk
> > problem. In several days before, the hd speed of this server become a
> > little slow than usual. And yesterday, during some basic operations on
> > the server, such as "ls", "cp", "ifconfig". Suddenly, there appeared a
> > I/O error, then the server died.
> >
> > The RAID controller is:
> > 0000:02:03.0 RAID bus controller: LSI Logic / Symbios Logic MegaRAID
> > (rev 01)
> > And there are a RAID-5 logical disk via three 146G SCSI physical disks.
> >
> > After boot the server, we have entered the RAID controller BIOS, found
> > all the physical disks has been the "FAILED/OFFLINE" situation.  Then we
> > forced the status of HD to ONLINE status.
> > Then we boot the system via a KNOPPIX cd, and the partition table
> > detected by fdisk look likes:
> >
> > root@4[root <mailto:root@4[root>]# fdisk -l
> > Warning: ignoring extra data in partition table 5
> > Warning: ignoring extra data in partition table 5
> > Warning: ignoring extra data in partition table 5
> > Warning: invalid flag 0xffffaa0a of partition table 5 will be corrected
> > by w(rite)
> >
> > Disk /dev/sda: 293.6 GB, 293628542976 bytes
> > 255 heads, 63 sectors/track, 35698 cylinders
> > Units = cylinders of 16065 * 512 = 8225280 bytes
> >
> >    Device Boot      Start         End      Blocks   Id  System
> > /dev/sda1   *           1          13      104391   83  Linux
> > /dev/sda2              14         650     5116702+  83  Linux
> > /dev/sda3             651        1287     5116702+  83  Linux
> > /dev/sda4            1288       35698   276406357+   5  Extended
> > /dev/sda5   ?        1675        2071     3177836+  dd  Unknown
> >
> > In fact, there should be sda5 and sda6, i guess cylinder 1289-1675
> > should be the sda5 - swap partition. And the rest should be the sda6.
> >
> > The first three partitions could mounted successfully, and sda contained
> > NO data which should be the boot partition. But sda2 and sda3 which
> > should be root partition and /var partition contained NO data too, but
> > there are so many files in the folder "lost+found".
> >
> > We have used GUI parted, parted, testdisk, gpart and some other disk
> > detect tools to the scan the disk and try to find out the correct
> > partition tables, but no results.
> >
> > So i post this problem here to ask for some advices.
> > Could someone pick me up?
> >
> > Thanks in advance.
> >
> > Penghui Wang.
> > --
> > Wang Penghui
> > http://www.extmail.org
> > http://www.postfix.org.cn
>
>
> --
> David Gardner
> --
> gentoo-server@g.o mailing list
>
>


-- 
Wang Penghui
http://www.extmail.org
http://www.postfix.org.cn