On 03/28/2011 03:43 PM, Carsten Aulbert wrote:
Hi
On Monday 28 March 2011 15:36:32 Mahmoud Mubarak wrote:
I'm using Ubuntu 10.10, kernel 2.6.35-27-server. Our network has over 1
Thousand nodes. When I tried to ssh to ~800 nodes, I got this error
message "No buffer space available" after executing ssh to ~600 nodes of
800. I'm just executing hostname command on each machine. How can I
handle this issue?
You might run into an issue with forking here (ulimits might help) or TCP
problems.
I suggest you use something like dsh et al. to manage large numbers of ssh
connections, e.g. dsh -aMF50 will run ssh with 50 hosts in parallel and will
still go through the large "default" list.
Does this help?
Cheers
Carsten
Thanks, Carsten.
well, Sorry for bad explain. I can elaborate: I had 20 groups with 40
nodes per group. I typed:
for i in group{01..40};do dsh -cg $i hostname>/dev/null;done
Then this error message appeared after group15. Any ideas about solving
this issue?
--
Mahmoud A. Mubarak
Bibliotheca Alexandrina
--
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
