[MITgcm-devel] mitgcm.org is very slow

[marshall-admin at techsquare.com]

hello jmc-

this should be all set now. forge ran out
of memory, invoked the oom-killer, and 
did not quite make it out cleanly. 

[greg]


Jean-Michel Campin <jmc at ocean.mit.edu> writes:

> Hi Greg,
>
> It's seems that a similar slow down is happening now:
> I tried some ssh to forge.csail.mit.edu that time-out, and when I managed to login,
> I see a load averaged > 16 (was at 9:00 am today, from "top" command), 
> with some busy httpd processes.
>
> You did something that helped last time (restarting some httpd related thing ?)
> may be it could be tried again ?
>
> Cheers,
> Jean-Michel
>
> On Tue, Jan 15, 2013 at 01:54:11PM -0500, Jean-Michel Campin wrote:
>> Hi Greg,
>> 
>> it looks like forge.csail.mit.edu (mitgcm.org) is busy (load averaged
>> from "top" command is currently over 35) and slowing down the
>> response from mitgcm.org
>> Could something been done ?
>> 
>> Thanks,
>> Jean-Michel
>> 
>> On Tue, Jan 15, 2013 at 06:04:28PM +0100, Martin Losch wrote:
>> > Hi there,
>> > 
>> > the access to mitgcm.org (both html and cvs) is very slow from here (awi.de). Do you also see this? www.mit.edu or other sites within the US seem normal to me.
>> > 
>> > Martin
>> > 
>> > 
>> > _______________________________________________
>> > MITgcm-devel mailing list
>> > MITgcm-devel at mitgcm.org
>> > http://mitgcm.org/mailman/listinfo/mitgcm-devel