SQL Error, Duplicate entry and HHH000315 in atlassian-crowd.log

Question

In production we have the following triple log entries:

2014-08-05 00:26:38,514 ajp-bio-8009-exec-12 WARN [engine.jdbc.spi.SqlExceptionHelper] SQL Error: 1062, SQLState: 23000
2014-08-05 00:26:38,515 ajp-bio-8009-exec-12 ERROR [engine.jdbc.spi.SqlExceptionHelper] Duplicate entry 'F0Ed288GGEXEw8iDZrzMVQ00' for key 'uk_token_id_hash'
2014-08-05 00:26:38,515 ajp-bio-8009-exec-12 ERROR [jdbc.batch.internal.BatchingBatch] HHH000315: Exception executing batch [could not perform addBatch]
2014-08-05 00:46:28,325 ajp-bio-8009-exec-11 WARN [engine.jdbc.spi.SqlExceptionHelper] SQL Error: 1062, SQLState: 23000
2014-08-05 00:46:28,326 ajp-bio-8009-exec-11 ERROR [engine.jdbc.spi.SqlExceptionHelper] Duplicate entry 'F0Ed288GGEXEw8iDZrzMVQ00' for key 'uk_token_id_hash'
2014-08-05 00:46:28,326 ajp-bio-8009-exec-11 ERROR [jdbc.batch.internal.BatchingBatch] HHH000315: Exception executing batch [could not perform addBatch]

As you can see, it occurs at different times. sometimes one time a day sometimes two or more times.

A look at the DB table cwd_token shows a valid token for one of our applications. Here is the row (name of application is replaced by xxxxxxxx):

id      directory_id entity_name random_number       identifier_hash          random_hash              created_date        last_accessed_date  last_accessed_time duration 
------- ------------ ----------- ------------------- ------------------------ ------------------------ ------------------- ------------------- ------------------ -------- 
1580324 -1           XXXXXXXXXXX 4671076784804601889 F0Ed288GGEXEw8iDZrzMVQ00 PAT0NGQA9ZjaUHkef0vuOA00 2014-08-05 08:11:43 2014-08-05 08:11:43 1407224594771      60

We are using latest Crowd-Version 2.7.2 Build 626.

We don't have any functional problems so far, but these log entries makes me wonder.

Giacomo Iacovelli · Answer

same problem with our setting: all latest versions: jira, confluence and crowd. some user got kicked out i.e. when they try to assign an issue - others when they do other actions. not really easy to find the root cause. JIRA shows me an incredible amount of user sessions and tomcat is spitting error on error.

unknown · Answer

Yes, that is what I mean. When I clear that setting and rebooted crowd: SQL problems disappeared and Crowd acted normally. No SSO tough, but authentication works from Jira / Confluence once again.

fyi: I am trying to get the SSO working between jira/confluence (all latest versions), but no luck yet :)

Jan Wittorf1 · Answer

Hi Mika,

if you mean the "SSO Domain" option in "General Options" of crowd adminstration: Yes, we have set this too. But we need to set it for cookie handling and security reasons.