rsh slow shutdown
Egan Ford
egan at sense.net
Fri Mar 14 11:21:38 PST 2003
rsh uses privileged ports (512-1023). So, there is a limited number of
them. Other processes use the privileged ports reducing the total number
available. rsh may also use 2 ports (one for stderr and one for stdout so
I've been told, but I have not seen it), and you'll also need one/processor,
not just one/node.
I recommend using ssh with non-privileged ports, then you can startup any
size job < 65536-1024 processors.
Other solutions exist using MPD and scalable rsh implementations (e.g.
pdsh).
Turn around and ask J.P., he knows about this.
> -----Original Message-----
> From: beowulf-admin at beowulf.org
> [mailto:beowulf-admin at beowulf.org] On Behalf Of rick bradshaw
> Sent: Thursday, March 13, 2003 9:04 AM
> To: beowulf at beowulf.org
> Subject: rsh slow shutdown
>
>
> I am wondering if anyone knows the solution to the following
> problem. I know first of all that rsh is not scalable but it should be
> able to hit 255 or so processes, but recently I received the following
> message when running 208 processes
>
> Timeout in waiting for processes to exit, 16 left. This may
> be due to a defective
> rsh program (Some versions of Kerberos rsh have been observed
> to have this
> problem).
> This is not a problem with P4 or MPICH but a problem with the
> operating
> environment. For many applications, this problem will only slow down
> process termination.
>
> I will also include the fact that I am running redhat 7.3 with rsh
> from the redhat rsh-0.17-5 rpm.
>
> Thanks in advance for the help.
>
> Rick Bradshaw
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org
> To change your subscription (digest mode or unsubscribe)
> visit http://www.beowulf.org/mailman/listinfo/beowulf
>
More information about the Beowulf
mailing list