[BGPCEP-380] BGP ingestion in Beryllium is too slow Created: 29/Jan/16 Updated: 03/Mar/19 Resolved: 05/Feb/16 |
|
| Status: | Resolved |
| Project: | bgpcep |
| Component/s: | BGP |
| Affects Version/s: | Bugzilla Migration |
| Fix Version/s: | Bugzilla Migration |
| Type: | Bug | ||
| Reporter: | Vratko Polak | Assignee: | Milos Fabian |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Operating System: All |
||
| External issue ID: | 5139 |
| Description |
|
While ingesting 1 million one-prefix BGP update messages takse 2-3 minutes with Lithium snapshot [0], corresponding Beryllium snapshot manages to ingest only around 300000 messages in 5 minutes [1]. There is no obvious reason for Beryllium to be slow this much, so this performance regression is a Bug. [0] https://jenkins.opendaylight.org/releng/view/bgpcep/job/bgpcep-csit-1node-periodic-bgp-ingest-only-stable-lithium/92/robot/bgpcep-bgp-ingest.txt/Singlepeer%20Prefixcount/Wait_For_Stable_Talking_Ipv4_Topology/ |
| Comments |
| Comment by Milos Fabian [ 02/Feb/16 ] |
|
The concern was addressed to the controller developers - https://lists.opendaylight.org/pipermail/controller-dev/2016-February/011531.html |
| Comment by Milos Fabian [ 03/Feb/16 ] |
|
Fixed by: https://git.opendaylight.org/gerrit/#/c/33970/ and https://git.opendaylight.org/gerrit/#/c/33971/ |
| Comment by Radovan Sajben [ 04/Feb/16 ] |
|
Test case reference: |
| Comment by Radovan Sajben [ 05/Feb/16 ] |
|
Updated test case references: bgpcep_bgp_ingest_txt_singlepeer_prefixcount_wait_for_stable_talking_ipv4_topology bgpcep_bgp_ingest_txt_singlepeer_changecount_wait_for_stable_talking_ipv4_topology bgpcep_bgp_ingest_txt_manypeers_prefixcount_wait_for_stable_talking_ipv4_topology bgpcep_bgp_ingest_txt_manypeers_changecount_wait_for_stable_talking_ipv4_topology |
| Comment by Vratko Polak [ 05/Feb/16 ] |
|
Beryllium run on Sandbox now fails [0] pointing to a search which lists only this Bug. But symptoms are different, the test sees zero prefixes, so a Bug different from this one should be open for that. |