[SATLUG] reisersfs vs other file systems

Geoff geofff at w5omr.shacknet.nu
Tue Jul 22 16:49:35 CDT 2008


Don Wright wrote:
> On Tue, 22 Jul 2008 14:38:53 -0500 (CDT), Ed Coates <edcoates at nighthawk.dyndns.org> wrote:
>
>   
>> Hate to burst the bubble Geoff, but I just executed the following command:
>>
>> dd if=/dev/hda1 of=/dev/null bs=1024 count=1000000
>>
>> and populated your messages file with over 2400+ lines about dma and 
>> opcodes and such as this:
>>     
>
> Lovely! An easy test. Do the other drives pass that test?  

well... as I said, I didn't think that the 80gig drive had a problem...

I'm gonna check the rest  that are mounted...

/dev/sda3 on / type reiserfs (rw,acl,user_xattr)
/dev/sda1 on /boot type reiserfs (rw,acl,user_xattr)
/dev/sdb1 on /new80gig type ext3 (rw)
/dev/hda1 on /160gig type reiserfs (rw)  <-- (issues)
/dev/sdc1 on /80gig-3 type reiserfs (rw)
/dev/hdb1 on /80gigger type xfs (rw)

I am watching, with 'less +F /var/log/messages' and didn't see -any- dma
issues on hdb1

note the time stamps from the command line prompt...

root at w5omr  Tue Jul 22  16:39:07
  ~
 :> dd if=/dev/sda3 of=/dev/null bs=1024 count=100000
100000+0 records in
100000+0 records out
102400000 bytes (102 MB) copied, 3.57352 s, 28.7 MB/s

  root at w5omr  Tue Jul 22  16:42:44
  ~
 :> dd if=/dev/sdb1 of=/dev/null bs=1024 count=100000
100000+0 records in
100000+0 records out
102400000 bytes (102 MB) copied, 2.8185 s, 36.3 MB/s

  root at w5omr  Tue Jul 22  16:43:31
  ~
 :> dd if=/dev/sdc1 of=/dev/null bs=1024 count=100000
100000+0 records in
100000+0 records out
102400000 bytes (102 MB) copied, 2.21278 s, 46.3 MB/s

  root at w5omr  Tue Jul 22  16:44:14
  ~
 :> dd if=/dev/sdc1 of=/dev/null bs=1024 count=100000
100000+0 records in
100000+0 records out
102400000 bytes (102 MB) copied, 0.520668 s, 197 MB/s

  root at w5omr  Tue Jul 22  16:46:38
  ~
 :> dd if=/dev/hda1 of=/dev/null bs=1024 count=100000
dd: reading `/dev/hda1': Input/output error
0+0 records in
0+0 records out
0 bytes (0 B) copied, 0 s, Infinity B/s

Now... let's copy some of /var/log/messages and look at -those- time stamps

Jul 22 16:45:31 w5omr dhcpd: DHCPREQUEST for 192.168.0.20 from
00:c0:a8:bf:7b:ad (gateway) via eth1
Jul 22 16:45:31 w5omr dhcpd: DHCPACK on 192.168.0.20 to
00:c0:a8:bf:7b:ad (gateway) via eth1
Jul 22 16:46:39 w5omr kernel: end_request: I/O error, dev hda, sector 63
Jul 22 16:46:39 w5omr kernel: printk: 55 messages suppressed.
Jul 22 16:46:39 w5omr kernel: Buffer I/O error on device hda1, logical
block 0
Jul 22 16:46:39 w5omr kernel: end_request: I/O error, dev hda, sector 71
Jul 22 16:46:39 w5omr kernel: Buffer I/O error on device hda1, logical
block 1
Jul 22 16:46:39 w5omr kernel: end_request: I/O error, dev hda, sector 79
Jul 22 16:46:39 w5omr kernel: Buffer I/O error on device hda1, logical
block 2
Jul 22 16:46:39 w5omr kernel: end_request: I/O error, dev hda, sector 87
Jul 22 16:46:39 w5omr kernel: Buffer I/O error on device hda1, logical
block 3
Jul 22 16:46:39 w5omr kernel: end_request: I/O error, dev hda, sector 63
Jul 22 16:46:39 w5omr kernel: Buffer I/O error on device hda1, logical
block 0


I'll do this again, after I've relocated the drive, just to see what
happens...

This is turning out to be a pretty interesting fact finding mission. 
Perhaps we can say, for sure, one way or another, that a 160gig drive,
or the -kind- of drive doesn't like to work on a Promise controller card.

-Geoff



More information about the SATLUG mailing list