Re: segmentation faults
Dirk Eddelbuettel <[email protected]>
| Newsgroups | gmane.linux.cluster.openmosix.general |
|---|---|
| Message-ID | <[email protected]> |
On 6 January 2006 at 13:53, Smemoe, Richard L. wrote:
| I have a openmosix cluster of 5 nodes running the 2.4.26 kernel and
| Redhat AS 3. We recently upgraded from an older version of RH and
| openmosix to this version. Until the upgrade, things were running
| smooth, but now binaries compiled using a fortran (Lahey Fortran)
| compiler first lock up the machine they are run on, then throw a
| segmentation fault and core dump after a while. When I disable the
| openmosix service, the binaries run fine. It doesn't seem to be
| specific to the fortran compiled programs though, since another program
| we also use, R, does the same thing.
|
| Any thoughts on how to begin to troubleshoot this?
I would try 2.4.27, a fairly time tested kernel. That would be as easy as
burning one clusterKnoppix cdrom, booting one box and boot the other using
the openmosixterminalserver. Then bring in your Fortran app via nfs, scp,
... and see how it does.
AFAIK the 'final' 2.4.27 patch dates from Nov 2004; clusterKnoppix and hence
Quantian had a slightly earlier one.
Another suggestion would be to update the Fortan to the version of RH you're
now running, ie to recompile them.
Hth, Dirk
--
Hell, there are no rules here - we're trying to accomplish something.
-- Thomas A. Edison
-------------------------------------------------------
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://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click