Actions
Story #2129
closed
As a user, I have docs on expected failover and heartbeat delay times for the /status/ API
Status:
CLOSED - WONTFIX
Priority:
Normal
Assignee:
-
Category:
-
Sprint/Milestone:
-
Start date:
Due date:
% Done:
0%
Estimated time:
Platform Release:
Groomed:
No
Sprint Candidate:
No
Tags:
Documentation, Pulp 2
Sprint:
Quarter:
Description
For pulp_workers, pulp_resource_manager, and pulp_celerybeat it takes between 300 and 390 seconds for a failure of a component to be noticed by the /status/ API. Also the pulp_celerybeat and pulp_resource_manager can be run in parallel and have their own failover times for one instance to recover for another. I'm not sure what those times are. The /status/ API is always expected to show up-to-the-second correctness for mongodb and Qpid.
These are nowhere in the docs, and they likely should be added to the /status/ page and possible a new page on it's own or the scaling guide.
Actions
.