News for hackers. Stuff that matters.

A comment on InfluxDB Clustering Design – Neither CP or AP. Parent comment

Re:InfluxDB Clustering Design – Neither CP or AP by lobster_johnson on Wednesday June 03, 2015 @08:51PM (#9656827)
I believe Raft is supposed to handle this; every modification is a log entry, and every recipient has to ack the log entry, a bit like two-phased commit.
[ Parent | on HN ]
Re:InfluxDB Clustering Design – Neither CP or AP by dcb18 on Wednesday June 03, 2015 @08:56PM (#9656839)
This is incorrect, only reads from the current Raft master are guaranteed to not be stale. In the case of InfluxDB, I think caching is safe because the shard metadata is immutable.
[ Parent | on HN ]
Re:InfluxDB Clustering Design – Neither CP or AP by schmichael on Wednesday June 03, 2015 @09:23PM (#9656907)
Where are you seeing that cluster metadata is immutable? I don't even know how that would work. Surely nodes, databases, shards, users, permissions, etc. all can change?
[ Parent | on HN ]
Re:InfluxDB Clustering Design – Neither CP or AP by dcb18 on Wednesday June 03, 2015 @09:37PM (#9656946)
Yep, you're right.
[ Parent | on HN ]
Re:InfluxDB Clustering Design – Neither CP or AP by schmichael on Wednesday June 03, 2015 @09:22PM (#9656906)
But the section I quoted said that local metadata queries go to the cache which is only updated "periodically."
[ Parent | on HN ]