search cancel

Unable to start discovery_server probe on child robot, JBDC connection exception error

book

Article ID: 134083

calendar_today

Updated On:

Products

DX Unified Infrastructure Management (Nimsoft / UIM)

Issue/Introduction

discovery_server probe will not start and throws JDBC connection error even though I am able to connect using the same Oracle information provided to the probe, the probe is unable to connect to CA UIM backend database.

Cause

- firewall(s) changes

Environment

Release: UIM 8.5.1 SP1

Component : UIMDIS

discovery_server 8.52

discovery_server probe is running on a different server than the core hub. 

hub version: 7.95

Resolution

First, follow the instructions to deploy the discovery_server probe to a child robot of the Primary (core) hub if this was not already completed.

https://docops.ca.com/ca-unified-infrastructure-management/8-5/en/installing-ca-uim/discover-systems-to-monitor/advanced-configuration/run-discovery_server-on-a-child-robot-of-the-primary-hub/

Confirm the database hostname and port number .

What happens when you run the command 'telnet localhost 1521' at the command prompt where the database is running?

Can you connect to the database outside of JDBC via a general client such as Toad?

Is the hostname correct in the listener.ora and tnsnames.ora files?

Try restarting the robot

   /etc/init.d/nimbus stop

   /etc/init.d/nimbus start

- Set the discovery_server probe loglevel to 5 and logsize to 100000, then,

- Deactivate it, wait till the port and pid disappear

- Activate it

- Open a support case and attach the entire discovery_server.log showing the startup and errors.

Test network/connectivity

- Check using ping from DS TO primary hostname and backend database

- tracert from DS TO primary hostname and backend database

- nslookup hostname and IP address (resolution) on both sides

DS takes the data connection/string FROM the data_engine, then connects on its own.

Check to make sure no local or remote firewall is interfering with the jdbc connection.

Check connectivity from the child robot TO the primary and the backend database using telnet if possible. If not use nc.

At first the nc command output seemed to be successful from the discovery_server host:

nc -zv <DB server hostname> 1521

Ncat: Version 7.50 ( https://nmap.org/ncat )

Ncat: Connected to 172.xx.x.xx:1521.

Ncat: 0 bytes sent, 0 bytes received in 0.01 seconds.

->>> But it was not. On the host where the jdbc connection test succeeds, the nc output should be:

nc -zv <DB server hostname> 1521

***Connection to <DB server hostname> 1521 port [tcp/xxxxx-xx] succeeded!***

That is the message/output that proves a connection is successful.

So check to make sure no one enabled the local firewall on that robot (check the firewall status/rules), just to cover that as well.

Then contact your network team to find out if they enabled any firewall in between or if the firewall is enabled on the database server or new rules were applied which blocks the child robot/ip from connecting to the DB server.