[GE users] Installing and Configuring SGE (as execute-host) on Diskless-Cluster Nodes

fx d.love at liverpool.ac.uk
Wed Apr 14 17:41:29 BST 2010


    [ The following text is in the "utf-8" character set. ]
    [ Your display is set for the "ISO-8859-10" character set.  ]
    [ Some characters may be displayed incorrectly. ]

zsomborg <zsombor at niif.hu> writes:

> We also have several diskless cluster, where the compute nodes get an
> IP address from DHCP (does not even have a DNS entry yet...),

I'm operating stateless (not diskless) nodes, but I'm not sure I
understand the situation.  Does that really mean DNS, as opposed to NSS
hosts?  How do the nodes get a name eventually, and why doesn't the DHCP
server give them one?  It seems odd for the nodes not to have fixed
names and addresses.

> and we don't want to list all the exec hosts manually. It would be
> very good if there would be a way to tell GridEngine that everybody in
> this subnet is an exec host.
>
> Or alternatively if we could tell GridEngine that everybody in this
> subnet is allowed to administer the cluster, then the exec nodes would
> be able to register themselves (however we would still need to find a
> way to add them automatically to the queue or the hostgroup).

You can arrange for the master to do arbitrary things when a
newly-booted host contacts a daemon on it, identifying it by the client
address, assuming the operation doesn't need to be authenticated.  (An
example is the way the pxeconfig system works with systemimager
installing stateful nodes to flip the PXE configuration after
installation.)

-- 
Dave Love
?E-Science?, Computing Services Department, University of Liverpool
AKA fx at gnu.org

------------------------------------------------------
http://gridengine.sunsource.net/ds/viewMessage.do?dsForumId=38&dsMessageId=253403

To unsubscribe from this discussion, e-mail: [users-unsubscribe at gridengine.sunsource.net].



More information about the gridengine-users mailing list