Back to bug 1369722
| Who | When | What | Removed | Added |
|---|---|---|---|---|
| John Schwarz | 2016-08-24 08:50:29 UTC | CC | jschwarz | |
| Miguel Angel Ajo | 2016-09-06 11:52:42 UTC | CC | majopela | |
| Red Hat Bugzilla Rules Engine | 2016-09-06 11:52:48 UTC | Target Release | --- | 10.0 (Newton) |
| Hynek Mlnarik | 2016-09-06 12:29:18 UTC | CC | hmlnarik | |
| Assaf Muller | 2016-09-20 12:55:53 UTC | Priority | unspecified | medium |
| Status | NEW | ASSIGNED | ||
| Target Release | 10.0 (Newton) | 7.0 (Kilo) | ||
| CC | ekuris | |||
| Link ID | OpenStack gerrit 192063 | |||
| Assignee | amuller | ihrachys | ||
| Flags | needinfo?(ekuris) | |||
| Assaf Muller | 2016-09-20 12:56:30 UTC | Target Milestone | --- | async |
| Flags | needinfo?(ekuris) | |||
| RHEL Program Management | 2016-09-20 14:07:44 UTC | Keywords | ZStream | |
| Ihar Hrachyshka | 2016-12-22 16:53:08 UTC | Status | ASSIGNED | ON_DEV |
| Ihar Hrachyshka | 2016-12-22 16:57:56 UTC | Status | ON_DEV | POST |
| Ihar Hrachyshka | 2017-01-05 20:25:39 UTC | Status | POST | MODIFIED |
| Fixed In Version | openstack-neutron-2015.1.4-12.el7ost | |||
| Ihar Hrachyshka | 2017-01-10 14:30:31 UTC | Doc Text | Cause: stale data database layer errors were not caught by server component, resulting in errors on port status update requests in particular racy scenarios. Consequence: sometimes, an L2 agent that attempted to update status for a port could catch an unexpected error from neutron-server, triggering costly full resync of port data. Fix: Now neutron-server correctly catches the previously missed error and repeats the failed operation. Result: L2 agent is no longer exposed to the unexpected error, and will not trigger full resync in this particular scenario. | |
| Doc Type | If docs needed, set a value | Bug Fix | ||
| errata-xmlrpc | 2017-02-03 16:01:40 UTC | Status | MODIFIED | ON_QA |
| Martin Lopes | 2017-02-13 04:10:13 UTC | CC | mlopes | |
| Doc Text | Cause: stale data database layer errors were not caught by server component, resulting in errors on port status update requests in particular racy scenarios. Consequence: sometimes, an L2 agent that attempted to update status for a port could catch an unexpected error from neutron-server, triggering costly full resync of port data. Fix: Now neutron-server correctly catches the previously missed error and repeats the failed operation. Result: L2 agent is no longer exposed to the unexpected error, and will not trigger full resync in this particular scenario. | Prior to this update, data database layer errors were not caught by the server component, resulting in errors during port status update requests (for particular race conditions). Consequently, on occasion, an L2 agent that attempted to update a port status could have caught an unexpected error from neutron-server, triggering a costly full resync of port data. With this update, neutron-server now correctly catches the previously missed error and repeats the failed operation. As a result, the L2 agent is no longer exposed to the unexpected error, and will not trigger full resync in this particular scenario. |
||
| Toni Freger | 2017-02-14 15:11:03 UTC | Status | ON_QA | VERIFIED |
| errata-xmlrpc | 2017-02-15 00:03:30 UTC | Status | VERIFIED | RELEASE_PENDING |
| errata-xmlrpc | 2017-02-15 13:50:35 UTC | Status | RELEASE_PENDING | CLOSED |
| Resolution | --- | ERRATA | ||
| Last Closed | 2017-02-15 08:50:35 UTC |
Back to bug 1369722