I think we've been getting lucky that all of our output formats are
"variable major" (blocking by variables). But there at least needs to be
some flexibility for handling parallel output formats
On Tue, Nov 12, 2013 at 4:09 PM, Kirk, Benjamin (JSC-EG311) <
benjamin.k...@nasa.gov> wrote:
> T
Thanks - the one size fits all approach for not serializing on processor 0 is
not clear to me - different formats might require a different approach, e.g.
blocking by DofObjects, blocking by variables? I assume you guys are most
interested in ExodusII?
I hope to have some more time to work on t
On Tue, Nov 12, 2013 at 4:06 PM, John Peterson wrote:
>
>
>
> On Tue, Nov 12, 2013 at 3:56 PM, Cody Permann wrote:
>
>> Here's a plot after applying John's "avoid serializing to all processes"
>> patch:
>>
>>
>> https://drive.google.com/file/d/0B8csupg5nQaaVVpOUFRHZ0FJX28/edit?usp=sharing
>>
>
>
On Tue, Nov 12, 2013 at 3:56 PM, Cody Permann wrote:
> Here's a plot after applying John's "avoid serializing to all processes"
> patch:
>
>
> https://drive.google.com/file/d/0B8csupg5nQaaVVpOUFRHZ0FJX28/edit?usp=sharing
>
And here's a branch with the patch Cody is referring to:
https://github.
Here's a plot after applying John's "avoid serializing to all processes"
patch:
https://drive.google.com/file/d/0B8csupg5nQaaVVpOUFRHZ0FJX28/edit?usp=sharing
Cody
On Mon, Nov 11, 2013 at 4:30 PM, Derek Gaston wrote:
> On Mon, Nov 11, 2013 at 5:23 PM, Cody Permann wrote:
>
>> This is a run wit
On Mon, Nov 11, 2013 at 5:23 PM, Cody Permann wrote:
> This is a run with 2 systems in it - the first one has 40 variables and
> totals about 25 million DoFs... the second one is two variables and comes
> out to a little over 1 million DoFs. This job is spread out across 160 MPI
> processes (and
On Mon, Nov 11, 2013 at 3:51 PM, Derek Gaston wrote:
> I've mentioned this a few times in passing - but now I have concrete
> evidence: Outputting the solution takes a TON of memory... way more than it
> should. This is something we've seen for a very long time - and it's one
> of the main reaso
