Details
Description
I feel this is very important for proper instrumentation and monitoring of our new functionality.
Given that forestDB pre-allocates its entire buffer cache, we don't yet have a way of exposing to the user how much of their cache is actually being used by live indexes. This will make it very challenging to know when more RAM is needed and/or whether the user can sustain more indexes within a given node running the indexer service.
I think this must be a blocker for GA if not for beta.
Attachments
For Gerrit Dashboard: MB-14592 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
50402,3 | MB-14592: Add fdb_get_buffer_cache_used() | sherlock | forestdb | Status: MERGED | +2 | +1 |
50430,1 | Merge remote-tracking branch 'origin/sherlock' | master | forestdb | Status: MERGED | +2 | +1 |
50634,2 | MB-14592 fdb: Add forestdb.BufferCacheUsed() API | unstable | indexing | Status: MERGED | +2 | +1 |
50635,2 | MB-14592 stats: Report used forestdb buffer cache | unstable | indexing | Status: MERGED | +2 | +1 |