[Glass] Why leave Admin GcGem and Reclaim GcGem running at all time?
Dale Henrichs via Glass
glass at lists.gemtalksystems.com
Mon Jun 22 11:02:08 PDT 2015
On 06/19/2015 04:01 PM, Mariano Martinez Peck via Glass wrote:
>
>
> On Fri, Jun 19, 2015 at 7:53 PM, James Foster
> <james.foster at gemtalksystems.com
> <mailto:james.foster at gemtalksystems.com>> wrote:
>
> > On Jun 19, 2015, at 3:47 PM, Mariano Martinez Peck <marianopeck at gmail.com
> <mailto:marianopeck at gmail.com>> wrote:
> >
> > I was just trying to see how I can minimize the memory consume
> while the stone/gems are not being too much used.
>
> I assume that the OS will swap out inactive processes, so that
> shouldn’t be much of a concern.
>
>
> I was also thinking about that. Do you know is seaside gems ever get
> to the idle state in which the OS would swap it's memory?
> On the other hand, I have the seaside maintenance vm running which
> expires sessions every 1 hour...so maybe that task makes all seaside
> gems none-idle ? mmmmm
>
I am not a student of Linux swapping algorithms, but I'm pretty sure
that the maintenance vm will not cause other seaside gems to "seem
non-idle".
The mainenance vm, will activate it's own memory pages and those pages
in the SPC that are needed as well, but the private memory of other gems
will not be affected ...
We really try to make sure that idle gems stay idle and will only wake
up when needed ... but there is at least one restless process that wakes
up every 30 seconds to facilitate the handling of SigAborts ... the
process itself doesn't touch much memory, but if a SigAbort is required
then the gem will also do an abort and page in a bit more memory, but an
abort is pretty inexpensive for a previously idle process ...
BTW, this little "sigabort process" would be a spot where you could add
the logic for running the periodic scavenge and mark sweep:)
Dale
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gemtalksystems.com/mailman/private/glass/attachments/20150622/f29f3a41/attachment.html>
More information about the Glass
mailing list