Oracle 11g listener causing failed tcp connection attempts on a non-RAC system

Environment: Oracle database, Oracle Linux 6.2

I was debugging a possible network problem on a Oracle Linux server with a fresh installation of Oracle 11gR2, when I noticed the following:

$ netstat -s

1178980 active connections openings
273496 passive connection openings
911657 failed connection attempts

The number of “failed connection attempts” increased by one for every 2 seconds. So, something was clearly trying to connect to something else and this failed for some reason. I didn’t find how I could further debug this problem with netstat. However, I had free access to the server, so I tried to shut down the Oracle services one by one to see if this would stop the growing number of failed connection attempts. When I stopped the Oracle listener service, it did…

Listener logging was on and there were a lot of warnings in the listener.log file:

WARNING: Subscription for node down event still pending

I did a search on My Oracle Support and I found more information in the MOS document with ID 372959.1. Apparently, the listener service tries to contact a RAC service, but, since this is a non-RAC installation, the connection fails. The solution is to add SUBSCRIBE_FOR_NODE_DOWN_EVENT_<listener name>=OFF to your listener.ora file, and then restart the listener. This should remove the warnings and stop the failed connection attempts.

This is how the listener.ora file now looks like for my listener with name “LISTENER”:

# listener.ora Network Configuration File: /u01/app/oracle/product/11.2.0/db_1/network/admin/listener.ora
# Generated by Oracle configuration tools.

      (ADDRESS = (PROTOCOL = TCP)(HOST = = 1521))

ADR_BASE_LISTENER = /u01/app/oracle

# see metalink note 372959.1