Thomas Jungblut
Thomas Jungblut
/test e2e-aws-ovn-etcd-scaling /test e2e-gcp-ovn-etcd-scaling /test e2e-azure-ovn-etcd-scaling /test e2e-vsphere-ovn-etcd-scaling
just to randomly add what I've found in https://github.com/etcd-io/etcd/issues/14501 - once we reconsider the balancer/resolver, we should also add some circuit breaking. There were some ideas by Easwar in https://github.com/grpc/grpc-go/issues/5672.
reading through #18175, maybe we can simply lift the limit by default? 5000 seems very smallish for the amount of throughput etcd can handle nowadays, especially when writing many small...
> In etcdserver.(*EtcdServer).LeaseRenew, I deleted ensureLeadership makes me wonder why it increases the CPU usage so much, it merely waits on read state notification. edit: `etcd_network_peer_sent_bytes_total are an order of...
do you mind to share a bit more about your setup? if you're running on AWS, on what instance type? you're also running five node etcd?
I have to bring this back from the dead again, we have a case where we ended up restoring a previously deleted member. I cobbled together a quick integration test...
I've been debugging this a bit now, the reason the removed member isn't added again is because of this condition in the raft apply loop: ``` if cc.NodeID == 0...