[Insight-developers] Parallel test failures
Bradley Lowekamp
blowekamp at mail.nih.gov
Thu Dec 17 11:32:52 EST 2009
I just committed the addition of the RUN_SERIAL test property to large IO tests. It should be compatible with the old CMake requirement.
Brad
On Nov 12, 2009, at 1:45 PM, Bill Hoffman wrote:
> Bill Lorensen wrote:
>> Does RUN_SERIAL require ctest 2.8? We only require 2.6 for itk currently.
>>
>
> It is just a test property and will be ignored by earlier versions of
> CMake. The set_tests_properties command is available in 2.6.
>
>
>
> -Bill
On Dec 17, 2009, at 10:23 AM, Lowekamp, Bradley (NIH/NLM/LHC) [C] wrote:
>
> On Dec 17, 2009, at 10:05 AM, Kevin H. Hobbs wrote:
>
>> Of the 8 hooperlab workstations submitting nightly builds of ITK
>> six are configured to use Nightly CMake and to run the tests in parallel.
>>
>> All of these show failures of the itkObjectFactoryTest2.
>
> I saw this one too, no idea off hand.
>
>>
>> Some of these show failures of the itkLargeImageWriteReadTest2.
>
> They appear to be timing out. I think that the issue is that multiple of the itkLargeImageWriteReadTest(s) are running at the same time and the system is running out of memory, swapping and timing out. I believe this issue was brought up before and there was some CMake option to run certain test sequentially, or serially or something?
>
>>
>> These failures do not happen on the 2 hooperlab workstations that use
>> ctest2.6-patch 4 and run the tests in serial.
>>
>> <signature.asc><ATT00001..txt>
>
> ========================================================
> Bradley Lowekamp
> Lockheed Martin Contractor for
> Office of High Performance Computing and Communications
> National Library of Medicine
> blowekamp at mail.nih.gov
>
>
> <ATT00001..txt>
========================================================
Bradley Lowekamp
Lockheed Martin Contractor for
Office of High Performance Computing and Communications
National Library of Medicine
blowekamp at mail.nih.gov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20091217/cef0dda7/attachment.htm>
More information about the Insight-developers
mailing list