Default retry handler causes infinite loop about etcd4j HOT 5 CLOSED

I do not understand your issue. -1 will retry indefinitely until etcd returns a valid response. Or is this issue an effect of issue #32 in which some valid responses are treated as errors?

from etcd4j.

Ok, I understand. The problem here is that if ETCD does not working, e.g. you use wrong url, this causes infinite loop.

I mean I tried to verify behavior how my component works when it's not possible to connect to etcd, e.g. because of network partition. My test terminated in infinite loop because of described problem.

Lets say you starting your server which needs to connect to remote etcd instance during init operation. The server will not work and you are not able to find out what happen as it just loops in infinite loop.

from etcd4j.

It indeed causes an infinite retry loop with the default retry handler. You can change to other retry behaviors. The log will tell what is going on while retrying:

[nioEventLoopGroup-2-11] DEBUG mousio.client.retry.RetryPolicy - Retry 10 to send command
[nioEventLoopGroup-2-12] DEBUG mousio.etcd4j.transport.EtcdNettyClient - Connection failed to https://127.0.0.1:4001

What do you propose? Use a different default RetryHandler which quits after some time?

from etcd4j.

I would suggest to break the loop if maxRetryCount either maxDelay exceeds defined limits.

from etcd4j.

Oh I see what you meant now. I overlooked the 1000ms max delay. Will fix it! Thanks for reporting!

from etcd4j.