Description
In addition to providing queryable stats (i.e. via a REST API), for debugging past issues it's vital that stats are periodically logged (in the same way that Couchbase Server does). Logging these to a file rather than keeping them in memory would allow for debugging issues even after a restart (or crash).
This should be local to a node, so that any issues are able to be investigated in isolation and without a dependancy on a larger cluster (for example, if a node had suspected hardware issues and was removed).
Attachments
Issue Links
- depends on
-
CM-115 Admin UI/Management Console for SGW (aka Mobile Manageability)
- To Do