[CONTROLLER-1542] Performance for MD-SAL data store Deletes reduced 100x in Boron Created: 29/Aug/16  Updated: 25/Jul/23  Resolved: 31/Aug/16

Status: Resolved
Project: controller
Component/s: mdsal
Affects Version/s: None
Fix Version/s: None

Type: Bug
Reporter: Jan Medved Assignee: Unassigned
Resolution: Cannot Reproduce Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Operating System: All
Platform: All

Attachments: Microsoft Word Data Store Performance Test.docx    
External issue ID: 6571

 Description   

There has been a significant performance regression for MD-SAL Deletes between Beryllium and Boron. Please see the attached document that compares data store performance between Beryllium and Boron using the dsbenchmark tests.

Performance for Deletes in Boron should be in the same ballpark as in Beryllium.

 Comments   
Comment by Jan Medved [ 29/Aug/16 ]

Attachment Data Store Performance Test.docx has been added with description: Boron MD-SAL Data Store performance report

Comment by Vratko Polak [ 30/Aug/16 ]

Current graphs do not show significant difference between Boron [0] and Beryllium [1] snapshots. I will run released Beryllium in Sandbox to see if the performance is different there.

[0] https://jenkins.opendaylight.org/releng/view/controller/job/controller-csit-1node-periodic-benchmark-only-boron/plot/getPlot?index=0&width=750&height=1000
[1] https://jenkins.opendaylight.org/releng/view/controller/job/controller-csit-1node-periodic-benchmark-only-beryllium/plot/getPlot?index=0&width=750&height=1000

Comment by Vratko Polak [ 30/Aug/16 ]

The benchmark job produces two pictures. One is "Performance by structure", the other one is "Performance by operations" (not sure how the two definitions differ).
Sandbox runs (Beryllium snapshot and Beryllium the original release) do not differ significantly [0] (per structure), [1] (per operations).

Is it possible that the alleged 100x decrease was caused by comparing the two different pictures?

Keeping this unconfirmed for now.

[0] https://jenkins.opendaylight.org/sandbox/job/controller-csit-1node-periodic-benchmark-only-beryllium/plot/getPlot?index=1&width=750&height=1000
[1] https://jenkins.opendaylight.org/sandbox/job/controller-csit-1node-periodic-benchmark-only-beryllium/plot/getPlot?index=0&width=750&height=1000

Comment by Vratko Polak [ 31/Aug/16 ]

This close to release (and release review), it is not desired to have a Bug which is both blocker and unconfirmed.

Setting to WORKSFORME, but please re-open this as soon as additional hints of a real bug appear.

Generated at Wed Feb 07 19:55:48 UTC 2024 using Jira 8.20.10#820010-sha1:ace47f9899e9ee25d7157d59aa17ab06aee30d3d.