Good day everyone,
I’m exploring WAMP/crossbario as a communication message bus for my project. I can’t find any documentation on how to upgrade the running crossbario instance - in a non-disruptive manner. This project is available for a long time, so I guess I am missing something.
Let me elaborate what I mean - when I have a multi instance service, e.g. ElasticSearch, I just perform a rolling upgrade over the cluster, one server instance a time. So when a particular server goes down, clients loose the connection and reconnect through load balancer to another instance. The downtime is very small.
With crossbario having just a single instance, on restart I will dump connections for all my clients at once and they will not be able to reconnect until the updated crossbario instance comes up.
Moreover, the router is the critical component in a system - without it nothing works, and having all my system being dependent on a single service instance (in fact the only part that I can scale out) feels uneasy.
So how do you guys manage corssbario updates in production?
Thanks in advance,