[Glass] Understanding temp memory used seaside gems and 'ps' output

[Dale Henrichs via Glass]

On 06/22/2015 10:02 AM, Mariano Martinez Peck wrote:
>
>
> On Mon, Jun 22, 2015 at 1:54 PM, Dale Henrichs via Glass 
> <glass at lists.gemtalksystems.com 
> <mailto:glass at lists.gemtalksystems.com>> wrote:
>
>     I assume that you have allocated a generaout amount of swap space?
>     With enough swap space, your system will continue running (slowly)
>     if you end up consuming all of available memory ...
>
>
>     On 06/19/2015 01:22 PM, Mariano Martinez Peck via Glass wrote:
>
>
> yes, I have plenty.
>
> $ cat /proc/meminfo | grep Swap
> SwapCached:         3632 kB
> SwapTotal:      17039352 kB
> SwapFree:       16996596 kB
>
> And yes, the system continues to work, but I noticed some random "slow 
> pauses" in the app of around 15 to 30 seconds and when I check "free 
> -m" I notice I am full with memory, so that's why I start to analyze 
> these things a bit.
>
Yeah paging can be expensive, but if you start seeing these kinds of 
pauses that can't be blamed on swapping, there are other things you 
might do ... the main one I'm thinking about is that Linux will 
prioritize disk reads over disk writes and that means that a commit 
(disk write for tranlog) can be delayed in favor of disk reads (disk 
swapping or SPC page reads). In this case you can move the tranlogs to a 
separate LUN reserved for your tranlogs ...

Dale
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gemtalksystems.com/mailman/private/glass/attachments/20150622/4b48978d/attachment-0001.html>