Monitor all NATS servers in a cluster

[glycerine]

So I've been running a little 2 node cluster, and watching one of the nodes with nats-top.

It was very surprising to see two out of the three clients disappear. But then I checked and realized that those two had simply failed over to the other gnatsd in the cluster. I had to verify this was the case by firing up a separate nats-top instance and pointing it at the other node in the cluster.

It would be of great utility and convenience to be able to monitor an entire cluster from one nats-top display.

Hence this suggestion: have one nats-top connect to all gnatsd in a cluster and provide information on all of them. I really want to see all that information in one place to not be panicked that some service has died unexpectedly.

[glycerine]

And thanks again for nats-top. It rocks!

[shinji62]

Hi,

Any update on that ? I mostly a 4 nodes cluster and have to connect to all of them each time, is not really convenient.

Thanks

[wallyqs]

@shinji62 think this would be a good addition, other servers could possibly be discovered via /routez/ if not set explicit as well. I will make updates soon to have this built into nats-top, thanks for the feedback 👍

[glycerine]

Ping. Hi @wallyqs! By chance is this in progress? If not, could you provide an outline of how it could be done? I don't really understand how /routez and related internal stuff works.

[oren-nonamesecurity]

Bump. Is there a plan or an alternative for that?

[jsirianni]

I was also surprised to see that this tool only returns data on the server it is connected to. It would be nice if could connect to all servers in the cluster.