Page MenuHomePhabricator

[EPIC] Elastic search service takes many hours to restart
Open, MediumPublic

Description

I believe the current restart takes 36 hours. It should be more than an order of magnitude faster.

Event Timeline

ksmith raised the priority of this task from to Medium.
ksmith updated the task description. (Show Details)
ksmith added a subscriber: ksmith.

Stakeholders: Cirrus operators and Cirrus Developers
Benefits: Faster restarts make operators less cranky and allow developers to have consistent restarts every week, month, whatever. More frequent pushing code.
Effort: Most effort is around pausing writes to Elasticsearch and verifying that we can upgrade to Elasticsearch 1.6.0.

Not sure if this ticket should be in the sprint? On its own there is nothing to do, its a tracking task for the other things that need to be done.

ksmith renamed this task from Elastic search service takes many hours to restart to [EPIC] Elastic search service takes many hours to restart.Jun 12 2015, 7:21 PM
ksmith edited projects, added Epic; removed Discovery-Search (Current work).
ksmith set Security to None.

Not sure if this ticket should be in the sprint? On its own there is nothing to do, its a tracking task for the other things that need to be done.

Good point. I took it out of the sprint, and marked it as an EPIC