Details
-
Task
-
Status: Closed
-
Major - P3
-
Resolution: Fixed
-
None
-
None
-
0.25
-
true
Description
Description
We would like to change the documentation about removing a shard to say that we do not expect change streams to be able to survive and be completely resumable when a shard is removed. As described in the linked tickets here, there is a possibility that the change streams will not be able to contact the removed shard for long enough.
Scope of changes (files that need work and how much)
- Remove Shards from Cluster: Should make a note here based on the description.
- removeShard cmd
- Change Streams: Could maybe add a Behavior section to this page documenting this.
- individual shell methods.
- Using this ticket to also flesh out the what is and isn't resumable based on changestream at collection/db/deployment level.
Read product documentation (design, spec, scope docs)
Read Changestream design docs to get an idea for any other pieces which might impact the additions to the docs.
MVP (work and date?)
Should take a few hours of writing effort between the three possible pages.
Attachments
Issue Links
- documents
-
SERVER-33778 Remove change_stream_remove_shard.js
-
- Closed
-
- related to
-
DOCS-11715 Docs for SERVER-34818: changeStream option startAtClusterTime should be a raw BSON Timestamp
-
- Closed
-