RE: kernel: aacraid: Host adapter reset request. SCSI hang ?

On Mon, 9 Jun 2003, Javier Rodriguez wrote:

>  
> With HyperThreading disabled, I was able to run with RedHat
> kernel-smp-2.4.20-13.9 for 4 days without a problem. The second server
> executing RedHat kernel-2.4.20-13.9 with HyperThreading disabled has been up
> for 6 days.

Hi all,
Here's a thought.

I'm going to try running with HT enabled, but setting the CPU affinity to
5, so binding to the 1st and 3rd CPUs (the non-virtual, if I
read dmesg correctly).

This is a dell PE2650 with HT enabled (2.4.20 + patches) which has had
problems with the aacraid timeouts.

I'll report back to the list on the results.

Regards
James Bourne

>  
> Javier
> 
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Javier Rodriguez
> Sent: Thursday, June 05, 2003 10:05 PM
> To: [email protected]; [email protected]
> Cc: [email protected]; [email protected]
> Subject: RE: kernel: aacraid: Host adapter reset request. SCSI hang ?
> 
> 
> I have the following update regarding testing with our two 2650 servers
>  
> - With HyperThreading enabled, I was able to execute RedHat kernel-2.4.20-9
> for 72 hours without a problem on both servers.
> - With HyperThreading disabled, I was able to execute RedHat
> kernel-2.4.20-13.9 for 49 hours without a problem on both servers.
>  
> This evening, I switched one of our 2650 servers to execute RedHat
> kernel-smp-2.4.20-13.9 with HyperThreading disabled. With HyperThreading
> enabled, this version would normally fail us in less then 24 hours (no
> longer than 48 hours), so well see what happens. After the reboot, I did not
> receive the 'Can't get RTC IO port" message in the dmesg log.
>  
> The other 2650 server is still executing RedHat kernel-2.4.20-13.9 with
> HyperThreading disabled.
>  
> Javier
>  
>  
> 
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of
> [email protected]
> Sent: Thursday, June 05, 2003 8:08 PM
> To: [email protected]
> Cc: [email protected]
> Subject: FW: kernel: aacraid: Host adapter reset request. SCSI hang ?
> 
> 
> Follow the path, seems we have problem with aacraid or scsi layer in smp
> kernels and possibly hyperthreading.
> Steve
> -----Original Message-----
> From: eric.rietjens-//[email protected] [mailto:[email protected]]
> Sent: Wednesday, June 04, 2003 1:38 AM
> To: Boley, Steve
> Subject: RE: kernel: aacraid: Host adapter reset request. SCSI hang ?
> 
> 
> 
> No openmanage software has been installed on our 2650 servers. 
> 
> The server with the SMP kernel but without hyperthreading is still running
> without problems (and it also produces the "Can't get RTC IO port" message).
> 
> 
> Eric. 
> 
> 
> 
> 	
> 
> 
> 
> 
> [email protected] 
> 
> 
> 03-06-2003 16:00 
> 
> 
>         
>         To:        Eric Rietjens/EHV/TASS/PHILIPS@EMEA3 
>         cc:         
>         Subject:        RE: kernel: aacraid: Host adapter reset request.
> SCSI hang ? 
> 
>         Classification:         
> 
> 
> 
> 
> 
> Do you have the openmanage loaded?  If so see if it does it without it
> loaded. 
> Steve 
> -----Original Message-----
> From: eric.rietjens-//[email protected] [mailto:[email protected]]
> Sent: Tuesday, June 03, 2003 6:58 AM
> To: linux-poweredge-Lp/cVzEoVyZskZv2Y/7f+AC/[email protected]
> Subject: Re: kernel: aacraid: Host adapter reset request. SCSI hang ?
> 
> 
> I have similar problems with three of our 2650 servers. If I use the SMP
> kernel (2.4.18-27.7.xsmp) I can reproduce the problem within 15 minutes. 
> If I use the non-SMP kernel (2.4.18-27-7.x) the problem does not occur. 
> 
> Today I ran some tests with the SMP kernel, but with hyperthreading
> disabled. Although not tested as extensively as  the non-SMP kernel, the
> problem 
> does not occur any more. So the SMP-kernel without hyperthreading on a 2650
> also seems to work. 
> 
> If I check the kernel log messages (dmesg) there is an error mesage when I
> use the SMP kernel: "Can't get RTC IO port". This message is not  generated
> when I use the non-SMP kernel. Could this be related with the RAID driver
> problems we are seeing? 
> 
> 
> Eric. 
> 
> 
> 
> 

-- 
James Bourne, Supervisor Data Centre Operations
Mount Royal College, Calgary, AB, CA
www.mtroyal.ab.ca

"There are only 10 types of people in this world: those who
understand binary and those who don't."

_______________________________________________
Linux-aacraid-devel mailing list
[email protected]
http://lists.us.dell.com/mailman/listinfo/linux-aacraid-devel
Please read the FAQ at http://lists.us.dell.com/faq or search the list archives at http://lists.us.dell.com/htdig/