Re: [Scst-devel] problem with lun scanning on RHEL 4

[Vladislav B. <vs...@vl...>]

Hi,

Sunil Patil, on 02/24/2010 07:57 PM wrote:
> Hi,
>  
> I am having a RHEL 4.8 machine (initiator). It has a qlogic hba qla2352 
> (2 Gb). I exported some vdisks to this machine and did following on the 
> machine,
>  
> # echo "scsi add-single-device <host> <channel> <target> <lun>" > 
> /proc/scsi/scsi
>  
> The scan failed. I saw following SCSI log messages:
>  
> Feb 24 16:06:52 imits047 kernel: scsi scan: INQUIRY to host 3 channel 0 
> id 2 lun 123
> Feb 24 16:06:52 imits047 kernel: scsi scan: 1st INQUIRY failed with code 
> 0x10000
> Feb 24 16:06:52 imits047 kernel: scsi: Unexpected response from host 3 
> channel 0 id 2 lun 123 while scanning, scan aborted
> Even full device scan failed with the same message. For full device 
> scan, REPORT_LUNS reported the correct number of luns including the one 
> that was newly exported. "issue-lip" (and also reboot) solves the 
> problem but we dont want to use that.
>  
> I observed that if I remove an already discovered device and re-add that 
> device then that succeeds. Its only a problem with newly  exported 
> devices for which scan fails.
>  
> I saw this problem on RHEL4.8 and RHEL 4.4 so far (it works on RHEL 5). 
> This also happens with a real array lun for ex, with Promise Array lun.
>  
> Has anybody seen this issue? Is it an issue with RHEL 4 or with qlogic 
> driver/firmware? Is there a fix/patch available for this?

This looks like well known problem in Linux kernel SCSI subsystem, which 
doesn't process online devices reconfiguration correctly. It's many 
times discussed in scst-devel and linux-scsi mailing lists.

Since you use RHEL, you can try to fill a bug/support request on RedHat.

Vlad