Some server nodes portal is not starting after restart
[DataSource] [down] [EP_PRT] [graceful restart] [LogConfigurator] [portal server] [restart] [server] [show on restart]
Symptom
After restart of cluster Environment, on some of the server nodes the Portal is not coming up.
On the screen you can see sometimes the following message:
“Application error occurred during request processing.
Details: java.lang.NullPointerException: null”
The defaultTrce files could show a varios of traces like:
java.sql.SQLException: Datasource name not found: Object not found in lookup of EP_PCD.
Other terms
DataSource, EP_PRT, server down, LogConfigurator
Reason and Prerequisites
You have a cluster with multiple server nodes. On start up of the portal each server node tries to register its portal loggers to the LogConfigurator Service.
Some times it is happens that one server node is trying to register loggers, during this registration some files are being opened for writing in the data base, while other server node is on the same phase already. This simultanious access to the data base is causing this issue.
Solution
A fix was provided on NW04s SP17
A workaround for this issue is to restart the problematic server nodes one after another.