[mvapich-discuss] [rsh] <defunct> ????

[Sayantan Sur]

Hello,

Thanks for contacting the group about this issue.

> I have a 8-node cluster, when I start my application, I saw
> many processes "[rsh] <defunct>", is this normal or I did
> something wrong ?

Can you please tell us if your application proceeds to completion?
Normally, mpirun_rsh uses `ssh' to exec jobs, unless specifically told
to use rsh (by using the -rsh flag). Do these "defunct" processes linger
around or do they dissapear when the job finishes?  As far as I know,
there shouldn't be any "defunct" processes left behind. 

Do you see this behavior when using "ssh" too?

> 
> I am using mvapich-0.9.5-119 and Mellanox IB Gold Distribution (IBGD) 
> v1.7.0 for Linux.

There is a much later version of IBGD (v1.8.1) which has several
bugfixes. Does this problem happen with that version too?

Thanks,
Sayantan.


-- 
http://www.cse.ohio-state.edu/~surs