[OmniOS-discuss] Chasing down scsi-related warnings

[Stephan Budach]

Hi all,

I am having trouble chasing down some network or drive-related errors on 
one of my OmniOS r018 boxes. It started by me noticing these errors in 
the syslog on one of my RSF-1 nodes. These are just a few, but I found 
almost every drive/LUN of that target node mentioned in the syslogd on 
the RSF-1 node:

Jul  3 15:51:01 zfsha01colt scsi: [ID 107833 kern.warning] WARNING: 
/scsi_vhci/disk at g600144f0564d504f4f4c3033534c3034 (sd4):
Jul  3 15:51:01 zfsha01colt     incomplete write- retrying
Jul  3 15:51:29 zfsha01colt scsi: [ID 107833 kern.warning] WARNING: 
/scsi_vhci/disk at g600144f0564d504f4f4c3033534c3035 (sd5):
Jul  3 15:51:29 zfsha01colt     incomplete write- retrying
Jul  3 15:55:25 zfsha01colt scsi: [ID 107833 kern.warning] WARNING: 
/scsi_vhci/disk at g600144f0564d504f4f4c3033534c3039 (sd6):
Jul  3 15:55:25 zfsha01colt     incomplete write- retrying
Jul  3 16:06:43 zfsha01colt scsi: [ID 107833 kern.warning] WARNING: 
/scsi_vhci/disk at g600144f0564d504f4f4c3033534c3135 (sd43):
Jul  3 16:06:43 zfsha01colt     incomplete write- retrying

Also, iostat -exM is showing HW errors for those LUNs, although I can't 
confirm that the actual drives are at fault on the iSCSI target, which 
is provided by another OmniOS box.

I then failed the zpools over from that target to the second HA node and 
the errors went along with it, so I am assuming that these errors are 
either network related to the storage node or maybe even 
drive/controller related to the storage node. However, I can't seem to 
pin point the problem. As these are only warnings, there is no visisble 
sign about any issue on the storage node, but nonetheless I'd like to 
know, what the underlying issue is.

Any ideas, anyone?

Thanks,
Stephan