Details
-
Improvement
-
Resolution: Won't Do
-
Major
-
3.1.3
Description
Currently, if the Admin credentials of the remote cluster in an XDCR replication change, there is no obvious feedback for this on the source side UI. Replication stalls with the standard errors:
2016-04-18 15:29:21 [Vb Rep] Error replicating vbucket 987. Please see logs for details.
|
2016-04-18 15:29:21 [Vb Rep] Error replicating vbucket 975. Please see logs for details.
|
2016-04-18 15:29:21 [Vb Rep] Error replicating vbucket 959. Please see logs for details.
|
...
|
and in the logs (e.g. error.log) it's pretty obvious what the problem is:
[ns_server:error,2016-04-18T15:30:21.112,ns_1@10.111.93.102:remote_clusters_info<0.1357.0>:remote_clusters_info:handle_info:447]Failed to grab remote bucket `default`: {error,rest_error,
|
<<"Authentication failed. Verify username and password. Got HTTP status 401 from REST call get to http://10.111.93.104:8091/pools. Body was: []">>,
|
{bad_status,401,<<>>}}
|
[ns_server:error,2016-04-18T15:30:51.117,ns_1@10.111.93.102:<0.6944.5>:remote_clusters_info:do_mk_json_get:1546]Request to http://Administrator:****@10.111.93.104:8091/pools failed:
|
{error,rest_error,
|
<<"Authentication failed. Verify username and password. Got HTTP status 401 from REST call get to http://10.111.93.104:8091/pools. Body was: []">>,
|
{bad_status,401,<<>>}}
|
However, we don't make this apparent to the user. As the error in the logs is pretty clear, it would make this much easier to diagnose if we exposed this in the UI.