You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The first server is stuck trying to elect a leader. Once the other nodes come online then a leader is elected and the nodes are created properly, but the first node never completes the initialization and needs to be restarted
It seems to me as though there shouldn't be a timeout there. if the node doesn't exist and we need to meta.CreateLocalNode, then we shouldn't timeout prematurely
The text was updated successfully, but these errors were encountered:
@jwilder I think this can be done in conjunction with the work on cluster joining #2966. Since @benbjohnson is going to be working on compression in the 0.9.2 release cycle, can you take a look at both this and the joining issue for 0.9.2?
On a new cluster, If i do the following,
start up one node, wait more then 5 seconds, start up a second node
then the first server will hit the timeout here:
https://github.com/influxdb/influxdb/blob/master/meta/store.go#L402
and will wait endlessly here:
https://github.com/influxdb/influxdb/blob/master/cmd/influxd/run/server.go#L298
The first server is stuck trying to elect a leader. Once the other nodes come online then a leader is elected and the nodes are created properly, but the first node never completes the initialization and needs to be restarted
It seems to me as though there shouldn't be a timeout there. if the node doesn't exist and we need to meta.CreateLocalNode, then we shouldn't timeout prematurely
The text was updated successfully, but these errors were encountered: