On 2019-08-27 8:19 PM, David Crayford wrote:
On 2019-08-27 8:13 PM, Andrew Rowley wrote:
FWIW - debugging performance of forked Unix process startup/overhead
is a
mess - we had one customer who was seeing terrible performance when
fork/execing tiny little shell processes that did practically
nothing. It
was only happening on one of their LPARs. The IBM Support center was
involved for a really long time, and I don't know if they ever really
figured out what was wrong.
It wasn't running under the bash shell was it? I noticed that bash
has very significant overhead compared to the regular shell.
I did some investigation using SMF data, and just wrote it up here:
https://www.blackhillsoftware.com/news/2019/08/27/comparing-bash-and-bin-sh-on-z-os/
Great write-up. Bash doesn't use local spawn so it always runs
sub-processes in child address spaces. I have reported this so Rocket
are aware
and have opened a ticket. I'm not sure if it will be addressed anytime
soon. Probably not.
Oops. Spoke too soon there! There is no "local" spawn as the job numbers
clearly show. How exactly are you deriving CPU time? Is it an
accumulation of SMF30CPT CPU times?
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
