[OPNFLWPLUG-729] Li plugin: Global peformance issues Created: 08/Jul/16 Updated: 27/Sep/21 Resolved: 08/Sep/16 |
|
| Status: | Resolved |
| Project: | OpenFlowPlugin |
| Component/s: | General |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | Bug | ||
| Reporter: | Luis Gomez | Assignee: | Tomas Slusny |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: All |
||
| Attachments: |
|
||||||||||||||||
| Issue Links: |
|
||||||||||||||||
| External issue ID: | 6178 | ||||||||||||||||
| Description |
|
As showed in all these tests: 1) Unstable stats collection: https://jenkins.opendaylight.org/releng/view/openflowplugin/job/openflowplugin-csit-1node-periodic-scale-stats-collection-daily-only-boron/ 2) Unstable RPC perf test (not all flows are added/removed): https://logs.opendaylight.org/releng/jenkins092/openflowplugin-csit-1node-periodic-rpc-time-measure-daily-only-boron 3) Unstable Switch scalability (100 to 400 switches): https://jenkins.opendaylight.org/releng/view/openflowplugin/job/openflowplugin-csit-1node-periodic-sw-scalability-daily-only-boron/ |
| Comments |
| Comment by Shuva Jyoti Kar [ 10/Jul/16 ] |
|
Can we disable the table-features collection (been turned on by default) for the li-plugin, and re-check the switch stability ? |
| Comment by Jozef Bacigal [ 11/Jul/16 ] |
|
Tomas Slusny is testing performance on deviceflowregistry, because we are creating cache before the role change and we a reading all config and opertional DS for the flows and it may cause slow donw the start up after device restart at all. Tomas is preparing a patch to let the deviceregistryflow run on another thread, we will see if the performance will go up. Patch is already done, but just testing it right now. |
| Comment by Jamo Luhrsen [ 11/Jul/16 ] |
|
(In reply to Luis Gomez from comment #0) for all of these jobs, if you look at the karaf.log produced you will see 2016-07-11 12:01:20,165 | WARN | entLoopGroup-5-3 | StatisticsGatheringUtils | 179 - org.opendaylight.openflowplugin.impl - 0.3.0.SNAPSHOT | Can't write to transaction: {} |
| Comment by Luis Gomez [ 19/Jul/16 ] |
|
Perf jobs look better this week except for switch scalability: |
| Comment by Anna Medvedova [ 15/Aug/16 ] |
| Comment by Tomas Slusny [ 22/Aug/16 ] |
|
Attaching document where I tried to compare ONOS Controller Performance Test Report by SDNCTC (one that Anna posted above) with our current OpenflowPlugin testing environment. |
| Comment by Tomas Slusny [ 22/Aug/16 ] |
|
Attachment tests.md has been added with description: SDNCTC ONOS compare with ODL OpenflowPlugin |
| Comment by Tomas Slusny [ 22/Aug/16 ] |
|
Attaching also HTML form of the document, also with some comparison images. |
| Comment by Tomas Slusny [ 22/Aug/16 ] |
|
Attachment tests.html has been added with description: SDNCTC ONOS compare with ODL OpenflowPlugin with images, HTML |
| Comment by Tomas Slusny [ 08/Sep/16 ] |
|
Switch scalability is only issue left, and Jozef is working on it in |