Asias He
Asias He
We saw similar problem here: https://github.com/scylladb/scylladb/issues/11354#issuecomment-1235150549
> But Scylla on this node was not able to restore connectivity w/ other nodes: this node == which node here? You mentioned node-2, which node you got the following...
Is this a new test? Did this test pass previously?
In the test performed here: https://github.com/scylladb/scylladb/issues/11609#issuecomment-1257148947 we stopped the nic on node4 at 2022-09-22T16:21:34+00:00 and resumed the nic at 2022-09-22T16:29:55+00:00. Did we perform any other operations, e.g., ip tables to...
@xemul Are you aware of any RPC issue that could make the send of the one way message hang forever in 5.1?
> @asias would `nodetool removenode` would need similar issue ? or there we don't relay on gossip, since the node is dead already ? (i.e. trying to think if there...
> > > @asias would `nodetool removenode` would need similar issue ? or there we don't relay on gossip, since the node is dead already ? (i.e. trying to think...
@avikivity Ping for merge. Useful for k8s deployment where ip changes.
> Sorry about the delayed review. @asias doesn't this conflict with repair-based-node-ops? RBNO allows streaming from the two surviving replicas, so it can decommission if a node is down. This...
It looks like the truncate failed not just because of the decommissioned node with ip 3.165. It needs 7nodes to be alive, but only 4 is alive. ``` < t:2022-03-26...