Skip to content
This repository was archived by the owner on Aug 23, 2023. It is now read-only.

index-rules.conf based pruning for bigtable #1108

Closed
Dieterbe opened this issue Oct 24, 2018 · 2 comments
Closed

index-rules.conf based pruning for bigtable #1108

Dieterbe opened this issue Oct 24, 2018 · 2 comments
Assignees
Milestone

Comments

@Dieterbe
Copy link
Contributor

#924 was rebased on master but I forgot to apply the new mechanism to the bigtable idx plugin

@Dieterbe Dieterbe added this to the 1.0 milestone Oct 24, 2018
@woodsaj
Copy link
Member

woodsaj commented Oct 24, 2018

@Dieterbe i idont think any changes are needed. The indexrules are loaded and handled by the memory-idx plugin, which both cassandra-idx and bigtable-idx use.

I dont see any changes in #924 made to the cassandra idx plugin

@woodsaj
Copy link
Member

woodsaj commented Oct 24, 2018

never mind, had another look and there are clearly changes needed.

@Dieterbe Dieterbe self-assigned this Oct 24, 2018
Dieterbe added a commit that referenced this issue Oct 26, 2018
1) both the at-runtime pruning of the memory index, as well as the
   on-startup initialisation of the cassandra index group metrics by
   nameWithTags, whereas the on-startup init of the bigtable index
   did it by name.  So even if all defs for a.b;foo=bar are stale,
   bigtable idx wouldn't prune it if there's e.g.
   a non stale a.b=foo=quux, in contrast with cassandra and memory
   index which consider these separate series for pruning.
   We fix it here, and also clarify some things a bit, making
   the cassandra and bigtable index implementation on par and easily
   diffable.

2) implement index-rules.conf based pruning for bigtable idx,
   rather than the old max-stale mechanism.
   fix #1108
@Dieterbe Dieterbe modified the milestones: 1.0, 0.11.0 Dec 12, 2018
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants