Hello there,
I'm using mesos / marathon to spawn my cluster, and I'm trying to
automatically disover nodes.
I found that mesos use a internal DNS to provide access to all nodes
extract:
dig test-rmq.marathon.mesos +short
172.16.17.3
172.16.6.3
but there is no reverse lookup for these ip in the internal mesos DNS.
So I found a way to start RMQ with rabbot@IP mode,
Working fine, my hosts start as
rab...@172.16.17.3
My issue, is when I start my second node, I can't resolve the DNS lookup
and fail.
Any idea how to prevent it ?
I'm think about a small script that will edit the config file with the
rights entries
Some log from the startup of the second node:
/rmq/test> 2018-05-06 20:49:21.193 [info] <0.193.0> Node database directory
at /var/lib/rabbitmq/mnesia/
rab...@172.16.6.3 is empty. Assuming we need to
join an existing cluster or initialise from scratch...
/rmq/test> 2018-05-06 20:49:21.193 [info] <0.193.0> Configured peer
discovery backend: rabbit_peer_discovery_dns
/rmq/test> 2018-05-06 20:49:21.193 [info] <0.193.0> Will try to lock with
peer discovery backend rabbit_peer_discovery_dns
/rmq/test> 2018-05-06 20:49:21.193 [info] <0.193.0> Peer discovery backend
does not support locking, falling back to randomized delay
/rmq/test> 2018-05-06 20:49:21.193 [info] <0.193.0> Peer discovery backend
rabbit_peer_discovery_dns does not support registration, skipping
randomized startup delay.
/rmq/test> 2018-05-06 20:49:21.202 [info] <0.193.0> Addresses discovered
via AAAA records of test-rmq.marathon.mesos:
/rmq/test> 2018-05-06 20:49:21.196 [info] <0.193.0> Addresses discovered
via A records of test-rmq.marathon.mesos: 172.16.17.3
/rmq/test> 2018-05-06 20:49:21.201 [error] <0.193.0> Reverse DNS lookup for
address 172.16.17.3 failed: nxdomain
/rmq/test> 2018-05-06 20:49:21.202 [info] <0.193.0> All discovered existing
cluster peers:
/rmq/test> 2018-05-06 20:49:21.202 [info] <0.193.0> Discovered no peer
nodes to cluster with
/rmq/test> 2018-05-06 20:49:21.204 [info] <0.33.0> Application mnesia
exited with reason: stopped
/rmq/test> 2018-05-06 20:49:21.321 [info] <0.33.0> Application mnesia
started on node '
rab...@172.16.6.3'
/rmq/test> 2018-05-06 20:49:22.999 [info] <0.193.0> Peer discovery backend
rabbit_peer_discovery_dns does not support registration, skipping
registration.
/rmq/test> 2018-05-06 20:49:23.075 [info] <0.193.0> message_store upgrades:
No durable queues found. Skipping message store migration
/rmq/test> 2018-05-06 20:49:23.075 [info] <0.193.0> message_store upgrades:
Removing the old message store data
/rmq/test> 2018-05-06 20:49:23.143 [info] <0.193.0> Creating user 'stagging'
Thank you