A fix for this issue is available in Confluence Data Center 9.1.0.
Upgrade now or check out the Release Notes to see what other issues are resolved.

We do see similar error while deleting space and space cannot be deleted in confluence server 7.13.7

2022-08-30 04:34:16,383 ERROR [Long running task: Space removal long running task] [persistence.dao.hibernate.HibernateLabelDao] lambda$findByLabel$0 There is more than one label named 'favourite' in the 'my' namespace.
 -- space: 204210206 | url: /spaces/doremovespace.action | traceId: 66d594bb93de1e11 | userName: xxxxxxxx| referer: https://XXXXXXXXXXXXX/spaces/removespace.action?key=ABC| action: doremovespace

Version 7.13.0 server edition is also effected by this problem. I get a lot of error messages in the atlassion-confluence.log like this:

// 2022-05-18 12:31:52,764 ERROR [http-nio-8090-exec-27] [persistence.dao.hibernate.HibernateLabelDao] lambda$findByLabel$0 There is more than one label named 'win10_iot' in the 'global' namespace.
 -- url: /confluence/admin/flushcontentindexqueue.action | traceId: e6032cda0391866b | userName: bernhard | action: flushcontentindexqueue 

How can I fix this?

also in server 7.4.5
we got 2663 lines of the table LABEL with the NAME = "favourite" and the NAMESPACE = "my"
is this really an error or a false positive?

We are also expiring this issue in Server 7.4.8.

021-09-01 14:56:32,184 ERROR [Caesium-1-2] [persistence.dao.hibernate.HibernateLabelDao] lambda$findByLabel$0 There is more than one label named 'xxxxxxxxx' in the 'global' namespace.

If I remember correctly (it's been a while), I looked at the source code (I worked for a client with a corresponding licence) and the problem was that the method for adding label was not thread safe.

I do not know if the implementation has changed. Back then it was possible when the server was under heavy load to add the same tag multiple times. If you look at the code the part should be not hard to find.

Hi Everyone, I was not able to replicate this issue (in our newest Confluence to date, 6.15.7) by attempting to add a new label in parallel, one from REST API and another from the UI. Is it only replicable from the Service Rocket Scaffolding plugin? Please let us know here if you have more information!

The versions 6.8.3 and 6.13.4 are also affected by this problem.

we are getting 100's of there messages where 'label-name' is one of about 10 different labels. 

2019-01-30 10:38:22,418 ERROR [http-nio-8089-exec-454] [persistence.dao.hibernate.HibernateLabelDao] lambda$findByLabel$0 There is more than one label named 'label-name' in the 'global' namespace.

I've seen a strange behavior related to this bug. When adding a label to a new page, if I type in the name of a label with the same name as an existing label but have different capitalization (e.g. "Star" instead of "star") and press RETURN, the label added is displayed as "star", but a search for label "star" doesn't pull the recently added page. Going back to the page, removing the label and typing in "star" or selecting from the existing labels will result in the page being properly found during a search.