Cause com.atlassian.crowd.exception.runtime.OperationFailedException: The following URL does not specify a valid Crowd User Management REST service: http://crowd.onlifehealth.com/crowd/rest/usermanagement/1/authentication?username=lance_lyons at com.

[lance_lyons]

Hi,

We recently changed our crowd instance to be SSL through apache mod proxy / reverse proxy.

We also changed the crowd.propertiers of Jira, Confluence, etc,etc to reflect https://crowd.onlifehealth.com.

We still get the following error which makes it seem that I missed changing one of the http://crowd.onlifehealth.com to https://crowd.onlifehealth.com however I cant find it anywhere.

What am I missing?

Cause

com.atlassian.crowd.exception.runtime.OperationFailedException: The following URL does not specify a valid Crowd User Management REST service: http://crowd.onlifehealth.com/crowd/rest/usermanagement/1/authentication?username=lance_lyons
at com.atlassian.crowd.embedded.core.CrowdServiceImpl.convertOperationFailedException(CrowdServiceImpl.java:855)

caused by: com.atlassian.crowd.exception.InvalidCrowdServiceException: The following URL does not specify a valid Crowd User Management REST service: http://crowd.onlifehealth.com/crowd/rest/usermanagement/1/authentication?username=lance_lyons<br< a=""> /> at com.atlassian.crowd.integration.rest.service.RestExecutor$MethodExecutor.executeCrowdServiceMethod(RestExecutor.java:443)

[Ivar]

Guys, this is a shot in the dark. But going back a couple of years I experienced something similar. The connection towards Crowd and "implicitly" to AD servers timed out. This was intermittent but enough to be annoying. It turned out that I connected to a a cluster of AD servers which again Crowd wasn't always that happy to work with. Once I changed from the url ad.ourservers.com to the internal ip address for one of the ad-servers specifically, the problem disapperared.

I'm mentioning this based on @Michaels last comment. It might be way off. But still, my 2 cents.

[Michael Downey]

For what it's worth, I have a support case open on this for the last month and counting, and work continues. Essentially the REST call times out when doing certain types of searches. Very small-scoped searches work, but the normal ones used by JIRA, Confluence, Bamboo, etc., don't and just time out. This seems to only happen for us when we use our delegated LDAP directory, and not a test-purpose local Crowd-based directory.

[Andy Bare]

Getting the same error when trying to administer Confluence... any thoughts?

[Andy Bare]

Turns out I just needed to update the Confluence application IP address under the Jira User Server

https://answers.atlassian.com/questions/59766/confluence-jira-remote-user-directory-is-not-functional-during-authentication

[Michael Downey]

Did you ever find an answer to this? We're having a similar problem where our client apps (e.g. JIRA & Confluence) are having trouble synchronizing to Crowd. We see this same message (as well as a "search" service too) in both JIRA and Confluence logs. Crowd doesn't seem to report back on anything useful.

[taylor pakula]

We're using Jira as a user directory source in Confluence. Same error, different config, similar root cause:

After switching host names (might also be caused by switching URL in any fashion), Confluence kept erroring on password resets. Also noticed that Confluence wasn't syncing from Jira - we had just forgotten to update the Jira URL in Confluence's user directories.

[Ivar]

I've seen the same issue. It came when I changed Confluence to use https://crowd.mydomain.com/crowd instead of https://1.1.1.1/crowd. Pinging using the domain name indicates that DNS name cannot be found.

What happens if you try and ping the new https address?

[lance_lyons]

Also to note.

This only happens when I try to administer the site in Confluence and user directories in Jira. I am able to login with no problem.