Am 04.04.2014 um 12:33 schrieb Tina Friedrich:

> Not sure how well defined these are - some more than other. I can give you my 
> understanding of them and then people can jump in and tell me where I'm wrong 
> :)

NB: As all definitions describe a certain feature, a machine can even belong to 
more than one of these at the same time. 


> Master Host ->        the SGE master, i.e. the machine that runs
>               the 'qmaster' process; that's the process that
>               handles the scheduling etc. So that's the machine
>               that 'decides' where a job goes, and the machine
>               you connect to with things like 'qsub', 'qstat', ...

To add confusion: Master Host sometimes refers to the Compute Node/Cluster 
Node/exechost of a parallel job, where the job was started on, i.e. where the 
`mpiexec` was executed on. The other machines participating in this parallel 
job are called Slave Hosts here.

It depends on the context.

-- Reuti


> Submit Host ->        (that's a well defined one) - a host in the list of
>               'submit hosts'. Means that the qmaster has been told
>               that jobs can be submitted from these hosts. As in,
>               if you run 'qsub job.sh' from a submit host, the
>               qmaster will schedule 'job.sh' for execution. If you
>               are not on a submit host, the qmaster would tell
>               you 'this host is not a submit host' and reject the job.
> 
> Compute Node ->       a machine provided for, well, computing - as in, a
>               machine that is meant to offer CPU time for compute
>               tasks (not, say, as an infrastructure server)
> 
> Cluster Node ->       a compute node that is part of a cluster - i.e. runs
>               a service that makes a qmaster aware of it's existance
>               and allows said qmaster to farm jobs to it (in SGE
>               terms, and execution host - running an 'execd' process)
> 
>               I think 'compute node' and 'cluster node' are quite
>               often used as / understood as being the same.
>               
> Head Node ->  Term that's quite often used for the Master Host. I'm
>               not sure on this one, but I always understood this in
>               the context of the 'traditional' cluster architecture
>               where your nodes are on a completely private network,
>               with one server - the head node - being 'in charge' of
>               the lot, often providing things like DHCP, NFS, ... to
>               the cluster nodes (and the qmaster service).
> 
> Login Node ->         That one, I'm not sure. I think I've heard this used
>               both for cluster nodes that allow interactive jobs,     
>               and for the submit host(s) in setups where only very
>               few hosts are allowed to submit (and you have to log in
>               to a designated machine to submit a job). And tons of
>               other things. Basically, a node that allows user logins.
> 
> As I've said before, only my 2c...
> 
> Tina
> 
> On 04/04/14 04:55, VG wrote:
>> I am working on SGE and I need some precise definition of some terms.
>> These are the terms I often hear about, but does not really understand them.
>> 
>> These are
>> 
>> |Head Node
>> Login Node
>> Compute Node
>> Submit Host
>> Cluster Nodes
>> Master Host
>> |
>> 
>> Any insight would be helpful
>> 
>> Thanks
>> 
>> Varun
>> 
>> 
>> 
>> _______________________________________________
>> users mailing list
>> [email protected]
>> https://gridengine.org/mailman/listinfo/users
>> 
> 
> 
> -- 
> Tina Friedrich, Computer Systems Administrator, Diamond Light Source Ltd
> Diamond House, Harwell Science and Innovation Campus - 01235 77 8442
> 
> -- 
> This e-mail and any attachments may contain confidential, copyright and or 
> privileged material, and are for the use of the intended addressee only. If 
> you are not the intended addressee or an authorised recipient of the 
> addressee please notify us of receipt by returning the e-mail and do not use, 
> copy, retain, distribute or disclose the information in or attached to the 
> e-mail.
> Any opinions expressed within this e-mail are those of the individual and not 
> necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot 
> guarantee that this e-mail or any attachments are free from viruses and we 
> cannot accept liability for any damage which you may sustain as a result of 
> software viruses which may be transmitted in or with the message.
> Diamond Light Source Limited (company no. 4375679). Registered in England and 
> Wales with its registered office at Diamond House, Harwell Science and 
> Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom
> 
> 
> 
> _______________________________________________
> users mailing list
> [email protected]
> https://gridengine.org/mailman/listinfo/users


_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to