TCLUG Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCLUG:14090] Disk I/O

Followup question. I took at look at /var/log/messages after doing the
tests:

===
Feb 25 19:56:11 dusk kernel: scsi : aborting command due to timeout : pid
513929, scsi0, channel 0, id 2, lun 0 0x08 0c 65 5f 80 00  
Feb 25 19:56:11 dusk kernel: scsi0: Aborting CCB #513941 to Target 2 
Feb 25 19:56:11 dusk kernel: scsi : aborting command due to timeout : pid
513930, scsi0, channel 0, id 2, lun 0 0x08 0c 65 df 80 00  
Feb 25 19:56:11 dusk kernel: scsi0: Aborting CCB #513942 to Target 2 
Feb 25 19:56:11 dusk kernel: scsi : aborting command due to timeout : pid
513931, scsi0, channel 0, id 2, lun 0 0x08 0c 66 5f 80 00  
Feb 25 19:56:11 dusk kernel: scsi0: Aborting CCB #513943 to Target 2 
Feb 25 19:56:11 dusk kernel: scsi0: CCB #513941 to Target 2 Aborted 
Feb 25 19:56:11 dusk kernel: scsi : aborting command due to timeout : pid
513932, scsi0, channel 0, id 2, lun 0 0x08 0c 66 df 48 00  
Feb 25 19:56:11 dusk kernel: scsi0: Aborting CCB #513944 to Target 2 
Feb 25 19:56:11 dusk kernel: scsi0: CCB #513943 to Target 2 Aborted 
Feb 25 19:56:11 dusk kernel: scsi0: CCB #513944 to Target 2 Aborted 
Feb 25 19:56:11 dusk kernel: scsi0: CCB #513942 to Target 2 Aborted 
Feb 25 19:56:11 dusk kernel: scsi0: CCB #513944 to Target 2 Aborted 
Feb 25 19:56:11 dusk kernel: scsi0: CCB #513942 to Target 2 Aborted
===

Anyone clueful on these matters?

Also, from /usr/src/linux/drivers/scsi/README.BusLogic:

===
More recently, BusLogic introduced the FlashPoint Host Adapters, which are
less costly and rely on the host CPU, rather than including an onboard
processor. Despite not having an onboard CPU, the FlashPoint Host Adapters
perform very well and have very low command latency.
===

D'oh!! Now I *really* have to get the Adaptec wide controller...


--
Christopher Reid Palmer : www.innerfireworks.com