feat(elasticsearch): use 7.10.2 cluster for local and staging #1192
Conversation
| @@ -43,6 +43,7 @@ app: | |||
| serviceAccountSecret: null | |||
| stackdriver: | |||
| enabled: false | |||
| elasticSearchHost: elasticsearch-1-master-hl.default.svc.cluster.local:9200 | |||
There was a problem hiding this comment.
I believe this should probably use the "normal" elasticsearch host rather than specifically the master headless one.
I believe in this context the hl version doesn't include the appropriate readiness checks (although we might have turned them off) and also we should probably be aiming at the non-master (coordinating only nodes) anyway (see like the second paragraph of here https://opster.com/guides/elasticsearch/high-availability/coordinating-only-dedicated-coordinating-node/)
Yes, but after all existing wikis have been re-indexed on the new Elasticsearch cluster. This can happen at the same time we begin reading from the new cluster. Otherwise, new wikis would have to wait until indexing is complete on all other wikis before they could start using search. |
|
You can specify a list: see wbstack/api#624 |
No description provided.