[ale] SCSI problems

[Jim Lynch]

Mike Kachline wrote:
> 
>         Guys,
> 
>         I'm getting a slew of errors to the tune of (see below .signature
> here). Has anyone seen these before? These occur on a fairly frequent
> basis, and I suspect are also related to the load encurrent on the drive.
> My questions are as follows:
>         1. Is "badblocks" really worth using? Does it cause the OS to
> avoid using "bad" parts of the drive?
>         2. Does this look like a dead drive to you too?
>         3. What is a good way to map SCSI ID's to /devs/sdbX's?
I can't help with the other questions, but as far as I can tell, Linux
assigns devices in the SCSI ID order.  If you have a 100Mb disk on SCSI
ID 4 and a 2 Gb disk on ID 6, then the 100 Mb disk would be /dev/sda and
the 2 Gb disk would be on /dev/sdb.  The partition table is used to
divide the disk up into /dev/sda[1-n] (or /dev/sdb[1-n]. 
Use fdisk -l to see what each device has on it.  For instance, here is
the output from fdisk -l on one of my systems.
bash-2.01# fdisk -l

Disk /dev/hda: 64 heads, 63 sectors, 1023 cylinders
Units = cylinders of 4032 * 512 bytes

   Device Boot   Begin    Start      End   Blocks   Id  System
/dev/hda1   *        1        1      260   524128+   6  DOS 16-bit >=32M
/dev/hda2          261      261     1023  1538208    5  Extended
/dev/hda3          729      729      748    40320   83  Linux native
/dev/hda4          749      749      768    40320   82  Linux swap
/dev/hda5          261      261      387   256000+   6  DOS 16-bit >=32M
/dev/hda6          388      388      728   687424+   6  DOS 16-bit >=32M
/dev/hda7          769      769      870   205600+  83  Linux native
/dev/hda8          871      871      946   153184+  83  Linux native
/dev/hda9          947      947     1023   155200+  83  Linux native

Disk /dev/hdb: 64 heads, 63 sectors, 525 cylinders
Units = cylinders of 4032 * 512 bytes

   Device Boot   Begin    Start      End   Blocks   Id  System
/dev/hdb1            2        2      524  1054368    5  Extended
/dev/hdb5            2        2      123   245920+   6  DOS 16-bit >=32M
/dev/hdb6          124      124      245   245920+   6  DOS 16-bit >=32M
/dev/hdb7          246      246      367   245920+   6  DOS 16-bit >=32M
/dev/hdb8          368      368      489   245920+   6  DOS 16-bit >=32M
/dev/hdb9          490      490      524    70528+   6  DOS 16-bit >=32M
bash-2.01# 

While these are IDE drives, not SCSI, they output would look similar. 
The hda and hdb would be replaced by sda and sdb.

Jim.

As far as your error messages go, I'd suspect a bad drive or bad/no
termination, but I'm not an expert.

> 
>                                                 Thanks,
>                                                         - Mike
> ====================================================================
> Michael Kachline - CS, Georgia Tech
> kachline at cc.gatech.edu
> http://brightstar.gt.ed.net/kachline/
> ====================================================================
> 
> <snip>

> <snip>

-- 
To see my .signature file, go to http://reality.sgi.com/jwl/signature/