Submitted patches

Page 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17  ...
Show: all, applied, pending, rejected
Patch Description Author Branch Submit Time Action Status Test Status
 ticket:1559 - WCPS: fix the fitToSampleSpace function correctly Summary: In WCPS, fitToSampleSpace function is used to fit the geo bounds (lower, upper) to a geo coordinate which corresponds to grid pixel. It needs to be fixed correctly so the bounding box of WCPS encoding will return correct values for xmax, ymin, then the pixel size of output is also correct. Test Plan: Run the test_all_wcst_import for the test coverage 10m and the output in WCPS result. Reviewers: vmerticariu, dmisev Differential Revision: http://codereview.rasdaman.org/D397   master   2017-05-15 16:48:05     APPLIED   SUCCESS 
 ticket:1506 - added test for directql data ingestion Summary: insertion of data using directql via a short script in test_select Test Plan: run ./test.sh in test_select Reviewers: dmisev Reviewed By: dmisev Differential Revision: http://codereview.rasdaman.org/D394   master   2017-05-15 13:03:08     APPLIED   SUCCESS 
 ticket:1547 - internal struct type sizes fix Summary: modified internal struct type size and offset computations, clarified some type recasting in ops.cc, and introduced tests. min_cells, max_cells, avg_cells, and add_cells should now work with user-defined struct types, and so should band-specific operations. Test Plan: ./test.sh in test_select Reviewers: dmisev Reviewed By: dmisev Differential Revision: http://codereview.rasdaman.org/D396   master   2017-05-15 12:55:23     APPLIED   SUCCESS 
 ticket:1552 - test if a string is json before attempting to parse it Summary: Currently, we try to parse json; if this fails, we print a warning and try to parse key/value then. This is not the best approach, as sometimes the format parameters might be json, but fail to parse because the json is invalid. In this case rasdaman should throw an exception instead of printing a warning. A better test for whether the format parameters are json, is to check if the first non-whitespace character is '{'; this patch implements this solution. Test Plan: Added a test to the conversion systemtest. Reviewers: bbell, bphamhuu Reviewed By: bphamhuu Differential Revision: http://codereview.rasdaman.org/D395   master   2017-05-13 17:20:38     APPLIED   SUCCESS 
 ticket:1553 - WCST_Import nilValues interval error for irregular time serie recipe Summary: The nilValuesi n irregular time serie recipe should be separated values, also if it has same values for floor() and ceil() it should be used the first one. Test Plan: Run the test for test wcst_import Reviewers: vmerticariu, dmisev Differential Revision: http://codereview.rasdaman.org/D393   master   2017-05-12 16:23:18     APPLIED   SUCCESS 
 ticket:1548 - / rasmgr auth error switched to warning and made clearer Summary: see title Test Plan: check log -- you may need to run make clean first Reviewers: dmisev Reviewed By: dmisev Differential Revision: http://codereview.rasdaman.org/D391   master   2017-05-11 11:42:15     APPLIED   SUCCESS 
 ticket:1506 - directql no longer fails to ingest data -- double free or corruption resolved Summary: the file contents pointer was being freed in two different locations of the code. Test Plan: directql your favourite file with insert or update Reviewers: dmisev Reviewed By: dmisev Differential Revision: http://codereview.rasdaman.org/D392   master   2017-05-11 11:40:03     APPLIED   SUCCESS 
 ticket:1537 - Release memory from Petascope after finishing response Summary: Petascope need at least 1.5 * size of result from Rasj (ideally is 2) + 300MB for each WCS (WCPS) request to get a big result (hundred MBs). However, the gabarge collector does not release the byte[] when it finished the response, so need to do it manually. Also the misleading error about the capabilitiy of servers is filtered more percisely. Test Plan: It is tested with a WCPS request which returns ~600 MB and Petascope needs to set to 1200 MB to work. After each request, the heap cache is reduced to < 100 MB. Reviewers: dmisev, vmerticariu Differential Revision: http://codereview.rasdaman.org/D387   master   2017-05-03 15:46:16     APPLIED   SUCCESS 
 ticket:1542 - check if rasdaman is properly started and stop it if it is not. Summary: when rasservers fail to initialize properly, rasmgr will keep trying to restart them (every 3 seconds by default), unnecessarily producing many log files. To fix this start_rasdaman.sh will try a simple select query after starting rasservers; if it fails to execute it means that no rasserver has probably started (select version() should always execute fine otherwise). In this case it shuts down rasmgr and hence prevents it going into the above loop of constantly restarting the rasservers. Test Plan: I've tested this manually by inserting a segfault into rasserver's initialization. It would be very hard to automate a test for this. Reviewers: vmerticariu, bphamhuu Reviewed By: vmerticariu, bphamhuu Differential Revision: http://codereview.rasdaman.org/D385   master   2017-05-02 18:57:51     APPLIED   SUCCESS 
 ticket:1537 - reduce rasj memory usage Summary: rasj uses 4-5x more memory than the array it eventually transports. The current patch reduces it to 3x, eventually with some more effort it can be brought down to 1-2x. Test Plan: manual profiling Reviewers: vmerticariu, bphamhuu Reviewed By: vmerticariu, bphamhuu Differential Revision: http://codereview.rasdaman.org/D386   master   2017-05-02 17:23:51     APPLIED   SUCCESS 

Submit a new patch

Select your locally generated patch file below for upload. Note that the patch can only be accepted if you agree to the Rasdaman Contribution Agreement by ticking the checkbox.

See How to submit patches for a step-by-step guide, in particular regarding the patch subject format. In order to ensure quick processing and to avoid rejection, you may want to make sure your patch adheres to the rasdaman code guide.


Select the code branch below to which you want to commit (help); if you are not sure what this means just leave the default (master).



I agree to the Rasdaman Contribution Agreement.