kriszos

this bug was probably introduced in 4.20 relevant thread https://discuss.linuxcontainers.org/t/lxc-move-error-websocket-bad-handshake/12754

root@n1:~# ss -tlpn | grep lxd LISTEN 0 4096 10.0.1.30:8443 0.0.0.0:* users:(("lxd",pid=1216,fd=20)) root@n1:~# ip addr 1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd...

Issue starts at 12ms in my production cluster, but in test environment that these logs are from it starts at around 30ms. Heartbeat errors shows only while moving instance with...

my test environment are virtual machines on the same host connected to single bridge. It is very unlikely that link is saturated.

Thanks, I am looking forward to resolution of this issue. Let me know if you need anything else.

That's a shame. Please do add information about latency limitation to documentation of clustering. Could you elaborate why latency is even relevant with lxd clusters. As far as I can...

@tomponline did some tests to check if there was any improvements after your changes. I tested on the same environment as before but upgraded to version 5.1. Unfortunately there was...