Bug 746899

Summary: On push operations, copyTrans runs too slowly
Product: [Retired] Zanata Reporter: David Mason <damason>
Component: PerformanceAssignee: Carlos Munoz <camunoz>
Status: CLOSED CURRENTRELEASE QA Contact: David Mason <damason>
Severity: low Docs Contact:
Priority: unspecified    
Version: 1.4.1CC: mkim, sflaniga, zanata-bugs
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: All   
Whiteboard:
Fixed In Version: 1.5.0 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2012-03-07 00:07:00 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 786670    
Attachments:
Description Flags
Proposed inversion of copytrans loops to iterate over TransUnits before locales none

Description David Mason 2011-10-18 07:54:13 UTC
Description of problem:
copyTrans runs too slowly

Version-Release number of selected component (if applicable):


How reproducible:

Steps to Reproduce:
1.Push from any client with copytrans enabled
2.Wait
  
Actual results:
It is slow

Expected results:
Slightly less slow

Additional info:

Comment 1 Carlos Munoz 2012-02-20 23:23:03 UTC
Refactored the copy trans service to perform less database accesses and avoid unnecessary invocations. The copy trans service is also being invoked directly after pushing a resource instead of using a Seam event.

Should be anywhere between 60-80% faster than before based on tests.

See:
https://github.com/zanata/zanata/commit/aa25f30d213dcefe0f70c9119c7202cdfe4be88c
https://github.com/zanata/zanata/commit/8b5200b649e7835f9e98851d3503445ed5a8c012
https://github.com/zanata/zanata/commit/9bb7ae05ff983943ed3c70a92627df54c74c06cd

Comment 2 David Mason 2012-02-23 06:31:29 UTC
Verified in 1.5

Push operations with copytrans now take up to ~30% less time with my test data.

Comment 3 David Mason 2012-03-01 03:12:24 UTC
Created attachment 566721 [details]
Proposed inversion of copytrans loops to iterate over TransUnits before locales

Master branch has diverged from this, it is included here for reference (would likely be a messy merge if applied).