bpsh and memory leak - wien
becker at scyld.com
Tue Oct 1 09:59:43 PDT 2002
On Tue, 1 Oct 2002, Florent Calvayrac wrote:
> We try to use WIEN97 on our Scyld beowulf cluster, and
> the following happens : the program lapw1 (more or less
> pure fortran 77), run interactively on the front node,
> happily grows to, say, 30Mo and then runs until completion.
> When run with bpsh on a remote node, the available memory
> just shrinks down until the system swaps to stall.
What is using the memory?
If you are running the compute nodes from a ramdisk, look for output
files in the ramdisk.
I was going to try out wien2k here, but apparently it's difficult to get
access to the source code.
> Any ideas ? Is it some kind of weird file access problem with NFS,
> wrong caching ? Free -t or meminfo seem to indicate that
> it it just the remote program growing forever.
You can use 'top' or 'ps' on the master to monitor memory usage of the
Donald Becker becker at scyld.com
Scyld Computing Corporation http://www.scyld.com
410 Severn Ave. Suite 210 Second Generation Beowulf Clusters
Annapolis MD 21403 410-990-9993
More information about the Beowulf