[torqueusers] ssh keys on compute nodes

"Mgr. Šimon Tóth" SimonT at mail.muni.cz
Wed Jun 16 10:52:34 MDT 2010


> Is it *really* required to have all compute nodes be able to log in to
> all other compute nodes in order to run jobs on nodes that do not use
> the head node as part of their job?
> Interactively, I can run a mpi job using mpirun that does not use the
> head node.  But if I specify the nodes I want to pick using PBS, it
> fails due to login problems.
> What am I missing here?

Well, if you have a multinode job the you always have a master node and 
slave nodes. Slave nodes talk only to the master node and the master 
node talks then back to the server.

The problem is the decision about what nodes in what roles will be used 
is done by the scheduler.

So unless you have a completely static environment and always run the 
same type of jobs, then yes it is *really* required.

Btw. just create the configuration on node node and then simply copy it 
to all others (plus use global instead of user configuration).

-- 
Mgr. Šimon Tóth

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 3366 bytes
Desc: S/MIME Cryptographic Signature
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20100616/37f50a5b/attachment-0001.bin 


More information about the torqueusers mailing list