Test Connection to LDAP in JTS setup wizard hangs for one hour
This happens in the JTS setup wizard when selecting custom setup. On the user registry setup page, we select LDAP, add all the configuration information and click Test Connection. There is no response for an hour and then we get error as below:
CRJAZ1560E: An error occurred while fetching connection status.
CRJAZ1560E: An error occurred while fetching connection status.
One answer
We checked in the WebSphere Application Server logs and found the below error:
[15/10/23 13:48:54:230 JST] 000002ca LdapConnectio I
com.ibm.ws.wim.adapter.ldap.LdapConnection DirContext
reCreateDirContext(String errorMessage) CWWIM4564I The user registry is
now connected to 'ldap://LDAP.XXXXXXXXX:389' LDAP Server.
[15/10/23 13:52:00:366 JST] 0000006d ThreadMonitor W WSVR0605W: Thread
"WebContainer : 44 @@ 13:39 jp11796 /jts/service/com.ibm.team.repository.
service.internal.IServerConfigurationRestService/configurationValidation"
(000002c2) has been active for 763574 milliseconds and may be hung. There
is/are 46 thread(s) in total in the server that may be hung.
The solution to this problem was to apply the below APAR. It is fixed in RTC 5.0.2 and there are iFixes in other versions such as 4.0.7, 5.0 and 5.0.1.
APAR PI26888
http://www.ibm.com/support/docview.wss?rs=180&uid=swg1PI26888
Prior to the fix in RTC the LDAP sync was not using the WAS classloader, but its own different classloader. When multiple requests are made for resources threads can hang. When we look at the logs you can see that more and more threads were hung up to 46 threads by the time we noticed. This makes WAS unresponsive.
If you have a symptom that looks similar to the second half of this error, this may be the cause. Note that by restarting the operating system (not WAS) you can sometimes temporarily overcome the issue.
[15/10/23 13:48:54:230 JST] 000002ca LdapConnectio I
com.ibm.ws.wim.adapter.ldap.LdapConnection DirContext
reCreateDirContext(String errorMessage) CWWIM4564I The user registry is
now connected to 'ldap://LDAP.XXXXXXXXX:389' LDAP Server.
[15/10/23 13:52:00:366 JST] 0000006d ThreadMonitor W WSVR0605W: Thread
"WebContainer : 44 @@ 13:39 jp11796 /jts/service/com.ibm.team.repository.
service.internal.IServerConfigurationRestService/configurationValidation"
(000002c2) has been active for 763574 milliseconds and may be hung. There
is/are 46 thread(s) in total in the server that may be hung.
The solution to this problem was to apply the below APAR. It is fixed in RTC 5.0.2 and there are iFixes in other versions such as 4.0.7, 5.0 and 5.0.1.
APAR PI26888
http://www.ibm.com/support/docview.wss?rs=180&uid=swg1PI26888
Prior to the fix in RTC the LDAP sync was not using the WAS classloader, but its own different classloader. When multiple requests are made for resources threads can hang. When we look at the logs you can see that more and more threads were hung up to 46 threads by the time we noticed. This makes WAS unresponsive.
If you have a symptom that looks similar to the second half of this error, this may be the cause. Note that by restarting the operating system (not WAS) you can sometimes temporarily overcome the issue.