Closed
Bug 1062342
Opened 10 years ago
Closed 9 years ago
update PHX Elasticsearch cluster to 1.2.x for Input and SUMO
Categories
(Infrastructure & Operations Graveyard :: WebOps: Other, task)
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: willkg, Assigned: cliang)
References
Details
(Whiteboard: [kanban:https://webops.kanbanize.com/ctrl_board/2/20] )
Elasticsearch made some big API changes between 0.90 and 1.0. All their libraries work with Elasticsearch 1.0+ and no longer work with 0.90. That makes development work difficult. Can we update the Elasticsearch cluster to 1.0 or 1.1 for Input?
Assignee | ||
Comment 1•10 years ago
|
||
If possible, I'd like to update to at least 1.2.x or 1.3.x, given the vulnerability disclosures around previous 1.x versions. [1] However, I'm not sure if disabling dynamic scripting is going to cause some serious repercussions. I can start a chain of bugs to upgrade the shared ES clusters but probably won't be able to give it much effort until next week. Will that time frame work for you? [1] http://www.elasticsearch.org/blog/scripting-security/
Reporter | ||
Comment 2•10 years ago
|
||
Most of us (Input, SUMO, Mozillians, etc) use ElasticUtils. The latest version of that goes to great lengths to support 0.90, 1.0 and 1.1 of Elasticsearch. It's pretty messy, but my theory was that everyone switches to that version, then we upgrade Elasticsearch to 1.1 and then we can upgrade ElasticUtils to a future version that doesn't support 0.90 anymore and thus is easier to maintain. I had difficulties making it also support Elasticsearch 1.2. I'll spend some more time on that and see if I can get over the hurdles I hit. Rob says Marketplace is using Elasticsearch 1.2 and they don't yet have a need for dynamic scripts. So I think disabling that is probably fine. We'd have to ask the owners of the other projects that use that cluster to know if they need it or not.
Assignee | ||
Updated•10 years ago
|
Assignee: server-ops-webops → cliang
Reporter | ||
Comment 3•10 years ago
|
||
I'm going to try to get ElasticUtils working with Elasticsearch 1.2 this week. I just haven't had time, yet. The reasons you gave in comment #1 to jump to 1.2 make a lot of sense to me. Can you give me a few more days before we go ahead with this?
Reporter | ||
Comment 4•10 years ago
|
||
I spent this morning working on ElasticUtils and got it working fine with Elasticsearch 1.2. Given that, I'm game for pushing forward to Elasticsearch 1.2.
Whiteboard: [kanban:https://kanbanize.com/ctrl_board/4/1187] → [kanban:https://webops.kanbanize.com/ctrl_board/2/20]
Assignee | ||
Updated•9 years ago
|
Summary: update Elasticsearch cluster to 1.0 or 1.1 for Input → update PHX Elasticsearch cluster to 1.2.x for Input
Reporter | ||
Comment 5•9 years ago
|
||
Rather than file another bug for SUMO, I'm adding SUMO to the summary. Input is currently using Elasticsearch 1.2.4 on the dev cluster which means our -stage environment doesn't match our -prod environment in some significant ways. That's a rough place to be. Definitely makes me nervous to the point where I've been pushing off all Elasticsearch-related development. Where are we at right now? Is there anything I can do to help move things along? Are there other bugs that are in progress that I don't know about and/or aren't referenced in this bug?
Summary: update PHX Elasticsearch cluster to 1.2.x for Input → update PHX Elasticsearch cluster to 1.2.x for Input and SUMO
Reporter | ||
Comment 6•9 years ago
|
||
Making this block on the actual webops work bug.
Assignee | ||
Updated•9 years ago
|
Flags: cab-review?
Assignee | ||
Updated•9 years ago
|
Flags: cab-review?
Assignee | ||
Updated•9 years ago
|
Assignee | ||
Comment 7•9 years ago
|
||
The PHX1 production ES cluster has been upgraded to version 1.2.4.
Status: NEW → RESOLVED
Closed: 9 years ago
Resolution: --- → FIXED
Updated•5 years ago
|
Product: Infrastructure & Operations → Infrastructure & Operations Graveyard
You need to log in
before you can comment on or make changes to this bug.
Description
•