Yes, This log entry is expected during the decommission and happens only once. But in my case I see these messages repeating several 1000 times a day and continuing. Even the restart of C* process didn't work.
On Sun, Oct 4, 2020 at 11:49 PM Erick Ramirez <erick.rami...@datastax.com> wrote: > yes correct, I gracefully decommissioned a few nodes from the cluster and >> I see the below log messages for all the decommissioned servers. >> >> INFO [GossipStage:51] 2020-10-05 05:54:10,429 Gossiper.java:1010 - >>> InetAddress /x.x.x.x is now DOWN >>> INFO [GossipStage:51] 2020-10-05 05:54:10,430 StorageService.java:1965 >>> - Removing tokens [100] for /x.x.x.x >> >> > That's expected everytime you decommission a node. The tokens the node > used to own get "removed" meaning the decommissioned node no longer owns > it. Those log entries are normal behaviour for a cluster. Cheers! >
