Error starting revad from sciencemesh-testing branch

[michielbdejong]

I think we need to add some config entries to make these errors go away:

2023-06-13 08:01:05.341 ERR ../reva/pkg/notification/notificationhelper/notificationhelper.go:77 > connecting to nats failed, notifications will be disabled error="connection to nats server at '' failed: nats: no servers available for connection" pid=301883 pkg=rhttp service=ocs
2023-06-13 08:01:05.342 INF ../reva/pkg/rhttp/rhttp.go:215 > http service enabled: ocs@/ocs pid=301883 pkg=rhttp
2023-06-13 08:01:05.343 INF ../reva/pkg/rgrpc/rgrpc.go:147 > grpc server listening at tcp:0.0.0.0:19000 pid=301883 pkg=rgrpc
2023-06-13 08:01:05.385 ERR ../reva/pkg/notification/notificationhelper/notificationhelper.go:77 > connecting to nats failed, notifications will be disabled error="connection to nats server at '' failed: nats: no servers available for connection" pid=301883 pkg=rhttp service=ocdav
2023-06-13 08:01:05.386 INF ../reva/pkg/rhttp/rhttp.go:215 > http service enabled: ocdav@/ pid=301883 pkg=rhttp
2023-06-13 08:01:05.386 ERR ../reva/cmd/revad/runtime/runtime.go:211 > error starting the http server error="http service sciencemesh could not be started,: sciencemesh: provider_domain is missing from configuration" pid=301883

[glpatcern]

For the notifications part, it's "benign" and @javfg already fixed the message AFAIU.

The last message is important: you need to specify provider_domain in the [http.services.sciencemesh] section (it was already mandatory, now reva won't start) and you can actually drop the invite_link template

[javfg]

@michielbdejong yes! I already have a patch ready so the connection to NATS is only performed if there is at least something configured for the notifications. That way the error will be an info log instead.

The failure to start should not be related to those errors, tho.

[michielbdejong]

And mesh_directory_url should be "https://sciencemesh.cesnet.cz/iop/meshdir"?