On Tue, Aug 22, 2017 at 7:55 PM, Mike Pontillo <[email protected]> wrote: > On Tue, Aug 22, 2017 at 6:40 AM, Flo <[email protected]> wrote: >> >> The firewall rules I requested are: >> allow from 10.0.252.0/24 to 10.1.0.61/32 tcp+udp/53, tcp/80, tcp/5240, >> tcp/5250-5259, tcp/8000, tcp/8888, udp/123 >> >> The rackd is able to connect to the regiond but the communication is >> not 100% functional. To me it looks like something is missing. >> The UI gets horribly slow after starting the remote rackd and getting >> the status of the remote rackd fails (UNKNOWN). > > > Can you attach the .log files in /var/log/maas on each controller? > > How many IP addresses are on your region controller? I've seen similar > issues in the past when the rack controller continues to try (and fail) to > connect to the region on filtered ports. Last week we landed a fix for a > portion of this issue[1] (which has not yet been released), which helps in > situations where the MAAS region has multiple secondary IP addresses on the > same subnet. Is it possible that the rack is trying to connect to the region > on an unreachable address?
It looks like that's exactly our problem. As we tried to use one Server as region-controller AND rack controller and another server as second rack controller (different fabric) the regiond announced all interfaces and IPs via RPC (http://regioncontroller/MAAS/rpc/ We will try to isolate the region controller and the "local" rack controller by using LXC. All the best, Florian -- Maas-devel mailing list [email protected] Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/maas-devel
