3 ms·
This would certainly help. High CPU was also an issue that we started to notice on 0.4.6, here with some of the NYC CoreOS guys, and that's been fixed. Chalk it
by barakm 12y ago
This would certainly help. High CPU was also an issue that we started to notice on 0.4.6, here with some of the NYC CoreOS guys, and that's been fixed. Chalk it up to completely redoing internal communication.
EDIT: Master election too :)
(Disclaimer: etcd dev here)
- eternalban 12y agoAre you guys still per spec RAFT or have you diverged at this point?
- philips 12y agoThe team has worked _very_ carefully to follow the raft state machine as described in the paper as close as possible. For example we have a set of tests[1] that takes possible problems outlined in the original paper and implements them as unit tests. [1] https://github.com/coreos/etcd/blob/master/raft/raft_paper_test.go https://github.com/coreos/etcd/blob/master/raft/raft_paper_t...
- bkeroack 12y agoKeep in mind that reads do not go through Raft by default, so it's possible to get stale data. They've added "consistent=true"/"quorum=true" URL parameters for GETs per https://github.com/coreos/etcd/issues/741 https://github.com/coreos/etcd/issues/741 as a workaround.