Re: Wierd slowdowns

Bruce Knox wrote:
> Below is the reference to the sys_time() patch.  
> While the sys_time patch can dramatically affect the speed of some
> programs, I would think you'd see this slowdown appear on different
> nodes as the processes ran over an extended period as with your
> application.  With one node behaving this way, I would be very
> suspicious of that node's hardware.  This could very well be a heat
> related problem.  Bruce

I never said it only happened on one node, I just gave node 32 as an 
example.  I think it's unlikely to be hardware related, because its 
happening on multiple nodes (I can't say for sure that it really happens 
on every single node, but I think so).  Also, a reboot of the node 
solves everything, if it were a hardware problem you wouldn't expect this.

One more interesting observation: whenever a node has "gone bad", i.e. 
jobs start running slowly, even turning openMosix off 
(/etc/init.d/openmosix stop) will not help.  After openMosix was turned 
off, processes started locally on that node will still run slowly.  This 
could be an indication this the problem has nothing to do with 
openMosix, but I really don't know how to test this further.

As for the time patch: do you mean the CONFIG_MOSIX_EXT_LOCALTIME option 
in the official 2.4.26 openmosix patch, or is there more?  Because this 
option is turned on in the kernels of our cluster.


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642