[mesa-users] Discrepancy in wd_cool_0.6M, version 5527

Ed Brown ebrown at pa.msu.edu
Tue Nov 19 15:51:02 EST 2013


More on this…

On Mac OS 10.8.5, SDK 20130408, 8 threads:  v5527 stopped with dt < min_timestep_limit after model 2450 (208 retrys, 59 backups).
v5596 stopped with dt < min_timestep_limit  after model 1830 (33 retrys, 0 backups).

Ed




On 19 Nov, 2013, at 14:08, Dean Townsley <Dean.M.Townsley at ua.edu> wrote:

> To add to Broxton's report...
> 
> For me with r5596 MESA/star stops at model 1851 when it reaches the max_number_retries of 400, instead of the dt limit like for r5527.
> 
> Dean
> 
> 
> On 11/19/2013 12:14 PM, Broxton Miles wrote:
>> Hey everyone,
>> 
>> I actually ran into this problem as well. I'm using our Linux cluster, which uses centros 5.5, with SDK version 20130408. 
>> 
>> So I checked out clean versions of 5596 and 5527 and reran the test suite problem and got the following results: v5527 failed at the same point as Dean and Naveh's runs, and version 5596 actually failed as well at model number 1851.
>> 
>> Broxton
>> 
>> 
>> On Tue, Nov 19, 2013 at 11:56 AM, Bill Paxton <paxton at kitp.ucsb.edu> wrote:
>> Hi Dean,
>> 
>> Thanks.  I'll ask you the same question that I posed to Naveh -- what happens when you try wd_cool_0.6M with the current version (5596)?   Is the problem still with us?
>> 
>> -Bill
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> On Nov 19, 2013, at 9:38 AM, Dean Townsley wrote:
>> 
>>> Hi Naveh, Bill,
>>> 
>>> Just as a point of information, I tried wd_cool_0.6M in MESA version 5527 and got the result that Naveh describes: it stops sometime between step 580 and 590 (it is set to print every 10 steps) with the dt < min_timestep_limit=1e-6.
>>> 
>>> I'm using SDK version 20130408 on an intel Linux machine, Ubuntu 13.10.
>>> 
>>> I know that doesn't help fix anything, but at least it confirms the issue on my architecture.
>>> 
>>> Dean
>>> 
>>> 
>>> On 11/18/2013 05:29 PM, Bill Paxton wrote:
>>>> Hi Naveh,
>>>> 
>>>> Here are some results for comparison.
>>>> 
>>>> When I ran the test suite before releasing 5527 (on my 12 core mac using gfortran from the MESA_SDK), wd_cool_0.6M finished okay after 7.5 minutes with 133 retries, 24 backups, and 3194 steps.
>>>> 
>>>> Similarly, before doing release 5271, wd_cool_0.6M finished okay after 6.9 minutes with 174 retries, 50 backups, and 3105 steps.
>>>> 
>>>> For the current release, 5596, wd_cool_0.6M finished okay after 4.89 minutes with 18 retries, 0 backups, and 2643 steps.
>>>> 
>>>> The various inlist_wd_cool_0.6M's are essentially identical except for the most recent in which I added this:
>>>> smooth_convective_bdy = .false.
>>>> 
>>>> Of course as Aaron mentioned, there may have been changes in controls.defaults too.
>>>> 
>>>> You should also check what you get with 5596 and let us know.
>>>> 
>>>> -Bill
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> On Nov 18, 2013, at 3:04 PM, Aaron Dotter wrote:
>>>> 
>>>>> Hi Naveh,
>>>>> 
>>>>> The first thing you ought to do is compare the default settings, see mesa/star/defaults/[controls.defaults,star_job.defaults], between 5271 and 5527.  Anything not set explicitly in the test_suite inlists will use the default values.  Any differences in the default values may be responsible for the results you've found, even though the WD inlists themselves are the same.  If not, it will require some deeper investigation.
>>>>> 
>>>>> Aaron
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> On Sun, Nov 17, 2013 at 5:04 AM, Naveh Levanon <naveh.levanon at gmail.com> wrote:
>>>>> I've noticed a discrepancy in running wd_cool_0.6M from the test suite.
>>>>> In my previous version, 5271, with no changes to test controls, the WD cools over 10Gyr (about 2800 steps) until reaching the Teff lower limit set for it.
>>>>> In version 5527, with the same controls, evolution stops abruptly after about 1.4Gyr (580 steps), with the timestep becoming too small. I couldn't find anything suspicious in the profiles to explain this behaviour.
>>>>> I removed and reinstalled both versions to ensure this wasn't my meddling with the test, and the discrepancy repeats itself.
>>>>> 
>>>>> Naveh
>>>>> 
>>>>> ------------------------------------------------------------------------------
>>>>> DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
>>>>> OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
>>>>> Free app hosting. Or install the open source package on any LAMP server.
>>>>> Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!
>>>>> http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>>>>> _______________________________________________
>>>>> mesa-users mailing list
>>>>> mesa-users at lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>>>>> 
>>>>> 
>>>>> ------------------------------------------------------------------------------
>>>>> Shape the Mobile Experience: Free Subscription
>>>>> Software experts and developers: Be at the forefront of tech innovation.
>>>>> Intel(R) Software Adrenaline delivers strategic insight and game-changing 
>>>>> conversations that shape the rapidly evolving mobile landscape. Sign up now. 
>>>>> http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk_______________________________________________
>>>>> mesa-users mailing list
>>>>> mesa-users at lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>>>> 
>>>> 
>>>> 
>>>> ------------------------------------------------------------------------------
>>>> Shape the Mobile Experience: Free Subscription
>>>> Software experts and developers: Be at the forefront of tech innovation.
>>>> Intel(R) Software Adrenaline delivers strategic insight and game-changing 
>>>> conversations that shape the rapidly evolving mobile landscape. Sign up now. 
>>>> 
>>>> http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk
>>>> 
>>>> 
>>>> _______________________________________________
>>>> mesa-users mailing list
>>>> 
>>>> mesa-users at lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>>> 
>> 
>> 
>> ------------------------------------------------------------------------------
>> Shape the Mobile Experience: Free Subscription
>> Software experts and developers: Be at the forefront of tech innovation.
>> Intel(R) Software Adrenaline delivers strategic insight and game-changing
>> conversations that shape the rapidly evolving mobile landscape. Sign up now.
>> http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk
>> _______________________________________________
>> mesa-users mailing list
>> mesa-users at lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/mesa-users
>> 
>> 
> 
> ------------------------------------------------------------------------------
> Shape the Mobile Experience: Free Subscription
> Software experts and developers: Be at the forefront of tech innovation.
> Intel(R) Software Adrenaline delivers strategic insight and game-changing 
> conversations that shape the rapidly evolving mobile landscape. Sign up now. 
> http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk_______________________________________________
> mesa-users mailing list
> mesa-users at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/mesa-users





More information about the Mesa-users mailing list