Batchtest fails

[gijzelaerr]

========== Running makems                                                
('========== $', 'makems /code/test/Batchtest/WSRT_makems.cfg')          
('========== $', 'mv WSRT.MS_p0 WSRT.MS')                                
('========== $', 'pyxis WSRT.MS ms.prep')                                
Traceback (most recent call last):                                       
  File "batch_test.py", line 90, in <module>                             
    run("pyxis WSRT.MS ms.prep"); #TODO: this is hacky, bug in CASAcore  
  File "batch_test.py", line 28, in run                                  
    raise RuntimeError("failed with exit code %x"%code);                 
RuntimeError: failed with exit code 1                                    

λ  /usr/bin/addbitflagcol WSRT.MS                                            
creating new column BITFLAG_ROW using an IncrementalStMan storage manager    
DATA column storage manager type is StandardStMan                            
BITFLAG column will use a default storage manager                            
Adding BITFLAG column with fixed shape: [4, 64]                              
[1]    25531 segmentation fault (core dumped)  /usr/bin/addbitflagcol WSRT.MS

[bennahugo]

Cannot reproduce @gijzelaerr. Maybe it is a KERN specific issue when you built Timba

However you did break pyxis in places :)

[bennahugo]

Ok no... can't python3ize this codebase

Running python 2 only I run into this issue as well

[o-smirnov]

But how is it even a python issue when addbitflagcol fails? It's a binary compiled from C++ (and casacore)....

[ratt-priv-ci]

I'm checking whether it is not a makems issue. I'm compiling and overriding the lofar makems files in usr/local to check

On Sun, Jan 13, 2019 at 3:17 PM Oleg Smirnov notifications@github.com wrote:

But how is it even a python issue when addbitflagcol fails? It's a binary compiled from C++ (and casacore)....

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/ska-sa/meqtrees-cattery/issues/88#issuecomment-453829008, or mute the thread https://github.com/notifications/unsubscribe-auth/ARFXpkQbzPYL7nSPo_7Ue7LfjHwMR6DXks5vCzHtgaJpZM4ZdugZ .

--

---

Benjamin Hugo

Junior Software Developer SARAO Black River Park, 2 Fir Street, Observatory, Cape Town, Western Cape, 7925 Contact: [+27] 0716293858 <+27%2071%20629%203858>

PhD. student, Radio Astronomy Techniques and Technologies, Department of Physics and Electronics, Rhodes University

Skype: benna.cn

[bennahugo]

Yup - it is LOFAR's fault! I've given ska-sa/makems some TLC and it compiles on bionic with the latest casacore. Building it from source in the dockerfile of cattery fixes the issue. @gijzelaerr you will have to do the same for your dockerfiles, unless you feel like adding it back into KERN is a good use of your time.

Baseline 576m, 6 ifrs: 0-4 1-5 2-6 3-7 4-8 5-9
Baseline 720m, 5 ifrs: 0-5 1-6 2-7 3-8 4-9
Baseline 864m, 4 ifrs: 0-6 1-7 2-8 3-9
Baseline 1008m, 3 ifrs: 0-7 1-8 2-9
Baseline 1152m, 2 ifrs: 0-8 1-9
Baseline 1296m, 3 ifrs: 0-9 10-12 11-13
Baseline 1332m, 2 ifrs: 0-10 9-12
Baseline 1404m, 2 ifrs: 0-11 9-13
00:00:00: Closing MS.
========== Verifying residual image
WSRT.MS.CORRECTED_DATA.channel.1ch.fits and /code/test/Batchtest/test-refresidual.fits differ by 1.84272e-05, this is within tolerance
========== Break out the bubbly, this hog is airborne!
Bye!

[gijzelaerr]

I think we should replace makems with writems which is now bundled with casacore?