Re: Customer box periodically becoming completely unresponsive -- we can't figure out why

[DrFranken <midrange@xxxxxxxxxxxx>]

Anything on the 'Front Panel' such as A6xx0255 ?  This smells of 
hardware that's coming and going.

Alternatively is it possible that severe disk I/O is swamping something? 
I was moving 220GB of large files to a 520 yesterday with FTP over GbE 
to a 6 drive RAID set and at times the partition would appear hung for 
30 or more seconds, including Telnet.


        - Larry "DrFranken" Bolhuis

www.frankeni.com
www.iDevCloud.com
www.iInTheCloud.com

On 10/2/2014 12:04 PM, James H. H. Lampert wrote:

> We have a customer whose Midrange box (a 525 running V7R1, with 4 LPARs,
> 55% system ASP used) has, recently, started becoming completely
> unresponsive. Even their Telnet is unresponsive. This particular time,
> it happened while I had a Telnet connection.
>
> We have an ILE-based server program, primarily using native record-level
> access, with some SQL done through CLI, and we have a web server that
> runs as a Tomcat context, which is a client to both our ILE-based server
> program and (via JDBC) to the SQL server. On this system, both our
> proprietary server and the Tomcat server run in their own subsystem.
>
> The system was rather quiet at the time it became unresponsive. There
> are no QSYSOPR messages going back before the system was abended and
> re-IPL'd, and the only history log entries I see from the approximate
> time when the system became unresponsive were perfectly routine ones,
> and one message -- after the system became unresponsive -- that was a
> CPF0A9B, "Probe failed for category QHTTP."
>
> I've been given one lead that I'm about to follow up on, namely a
> potential SQL problem (even though I see nothing in any QZDASOINIT
> joblogs), but does anybody else have any ideas of what could be causing
> this?
>
> --
> JHHL