> It looks like from the man page of sge_types that only k, K, m, M, g and G
> are accepted. Since we are just starting to see TB of RAM as a realistic
> possibility in systems, perhaps a feature request should be the inclusion of
> t and T in future releases?

I think so!! The type MEMORY has been associated with system memory,
but in our case, we use the memory type to relate to actual disk
storage on a node. I am sure other use cases exists besides ours.
Depending on how ambitious I am feeling, I may dive into GE source and
see if I can submit a patch file over to Son of Grid Engine.

> BTW, the lower case values are based on multiples of 1000 bytes whereas the
> upper case value is 1024 bytes.

Ah good to know! Looks like reading man pages might be helpful for me :-)

FYI: I ended up converting all files to G.

Thanks Ian,
-Adam

--
Adam Brenner
Computer Science, Undergraduate Student
Donald Bren School of Information and Computer Sciences

Research Computing Support
Office of Information Technology
http://www.oit.uci.edu/rcs/

University of California, Irvine
www.ics.uci.edu/~aebrenne/
[email protected]


On Mon, Apr 1, 2013 at 8:42 AM, Ian Kaufman <[email protected]> wrote:
> It looks like from the man page of sge_types that only k, K, m, M, g and G
> are accepted. Since we are just starting to see TB of RAM as a realistic
> possibility in systems, perhaps a feature request should be the inclusion of
> t and T in future releases?
>
> BTW, the lower case values are based on multiples of 1000 bytes whereas the
> upper case value is 1024 bytes.
>
> Ian
>
>
> On Sat, Mar 30, 2013 at 8:41 PM, Adam Brenner <[email protected]> wrote:
>>
>> Howdy GE Users,
>>
>> We are running GE 8.1.2. I have setup a consumable resource on our
>> cluster that will monitor available local storage on each of our
>> execution nodes/hosts. Our expected values can be in the range of GB
>> to TB.
>>
>> qconf -mc entry:
>>
>> #name               shortcut   type        relop   requestable
>> consumable default  urgency
>>
>> #------------------------------------------------------------------------------------------
>> scratch_size        scratch    MEMORY      <=      YES         YES
>>    0        0
>>
>> When running qconf -mattr on the specific node, we get the following
>> error.
>>
>> [root@hpc2 scratch-cron]# sh scratch-cron.sh -r
>> attribute "scratch_size" is not a memory value
>> FAILED: qconf -mattr exechost complex_values scratch_size=1.6T hpc2
>> produced exit code 1
>> [root@hpc2 scratch-cron]#
>> (converting 1.6T to 1638.4G works fine with the above command)
>>
>> Is it by design that memory values do not exceed GB? Should I convert
>> everything to GB values when using qconf -mattr?
>>
>> Thanks,
>> -Adam
>>
>> --
>> Adam Brenner
>> Computer Science, Undergraduate Student
>> Donald Bren School of Information and Computer Sciences
>>
>> Research Computing Support
>> Office of Information Technology
>> http://www.oit.uci.edu/rcs/
>>
>> University of California, Irvine
>> www.ics.uci.edu/~aebrenne/
>> [email protected]
>> _______________________________________________
>> users mailing list
>> [email protected]
>> https://gridengine.org/mailman/listinfo/users
>
>
>
>
> --
> Ian Kaufman
> Research Systems Administrator
> UC San Diego, Jacobs School of Engineering ikaufman AT ucsd DOT edu
>
> _______________________________________________
> 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