Use one wsgi app for service but one dbengine per worker
Bug #1089916 reported by
Therese McHale
This bug affects 1 person
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Glance |
Fix Released
|
Medium
|
Therese McHale | ||
Grizzly |
Fix Released
|
Medium
|
Therese McHale |
Bug Description
In Bug 1036193 multiple workers were broken (both api and registry) . This was because glance was using 1 dbengine per worker process and this was causing garbled database sessions. The fix which was implemented in late folsom changed from using 1 wsgi app for each service to using one wsgi app per worker process (and therefore we got 1 dbengine per worker). With one wsgi app per worker this causes the environment to be printed for each worker. An alternative approach is go back to one wsgi app but change to using a separate dbengine for each worker by creating the engine later. (This is what nova does also.) Including a patch for this.
Changed in glance: | |
milestone: | none → grizzly-2 |
status: | Fix Committed → Fix Released |
Changed in glance: | |
importance: | Undecided → Medium |
To post a comment you must log in.
Fix proposed to branch: master /review. openstack. org/18037
Review: https:/