Better error messages in ocmd flow

[michielbdejong]

As a first example:

ocm-invite-forward -idp cernbox.cern.ch -token 15319544-a2a0-42a6-bcc1-bf63de698a7d
error: code=CODE_INTERNAL msg="error forwarding invite" support_trace="79f38ef674f16ed11dd01ee896e0fa19"

Is it that an invite with that token has already been accepted? The invite store backend is not writable? The other reva instance is not reachable? The other reva instance returned an error code? If so which one?

[michielbdejong]

Another urgent one: https://github.com/cs3org/reva/blob/v1.19.0/internal/http/services/ocmd/send.go#L119 - see pondersource/oc-sciencemesh#22

[michielbdejong]

Working on this in the ocmd-error-messages branch

[michielbdejong]

Improved error message for the first one:

>> ocm-invite-forward -idp cernbox.cern.ch -token asdf
error: code=CODE_INTERNAL msg="error forwarding invite:json: error sending post request: Post \"http://127.0.0.1:19001/ocm/invites/accept\": dial tcp 127.0.0.1:19001: connect: connection refused" support_trace="1118eb2354c77504651e3d25f7896c22"

[michielbdejong]

A third one (from sciencemesh/oc-sciencemesh#4):

In addition, note the OK response to file Delete GRPC call, even though the nextcloud storage driver returned a HTTP 500 HTML error page because of the PHP error above

[michielbdejong]

Fourth one: in the ocm-test-suite setup, when looking up sharees in oc1.docker, I'm seeing a call to /cs3.gateway.v1beta1.GatewayAPI/FindAcceptedUsers but I can't see the details of it.

[michielbdejong]

>> recycle-purge
error: code=CODE_INTERNAL msg="error emptying recycle:Unexpected response code from EFSS API: 500" support_trace="124d5fc6e935886db6dae181b1fae9d1"