Meridian Startup

This section describes troubleshooting steps for problems encountered while Meridian starts.

Java errors

If you encounter a Java error when starting Meridian, you should verify that your Java version is compatible:

runjava -s

If you run systemctl start opennms with a version of Java that is not compatible, the output displays error messages:

opennms[80542]: Starting OpenNMS:
opennms[81178]: runjava: Error: configured JVM not found.
opennms[81178]: runjava: "/usr/lib/jvm/java-11-openjdk-11.0.17.0.8-1.el9_1.x86_64/bin/java" does not exist or is not executab>
opennms[81178]: runjava: Run "/opt/opennms/bin/runjava -s" to set up the java.conf file.
opennms[81178]: runjava:
opennms[81178]: runjava: You can install a JVM by downloading one from Oracle, or by running something like
opennms[81178]: runjava: 'apt-get install openjdk-11-jdk' or 'yum install java-11-openjdk-devel'.
opennms[80542]: OpenNMS was unable to connect to the 'opennms' database configured in opennms-datasources.xml.
opennms[80542]: failed
systemd[1]: opennms.service: Control process exited, code=exited, status=1/FAILURE

See Minimum System Requirements for software compatibility requirements, including supported Java versions.

Configuration errors

If you encounter a configuration error when starting Meridian, you may need to run the Meridian database installer. This updates the relational database schema to the latest version and applies any required changes to the configuration files:

install -dis

Running or re-running the database installer is generally safe, provided you stop Meridian first.

If you run systemctl start opennms and Meridian is not installed or up to date, the output displays error messages:

opennms[119337]: /opt/opennms/etc/configured does not exist.
opennms[119337]: You need to run the installer to set up the database.  In most
opennms[119337]: cases, it is enough to run:
opennms[119337]:   /opt/opennms/bin/install -dis
opennms[119337]: For details, see the install guide at:
opennms[119337]: http://www.opennms.org/index.php/QuickStart#Initialize_OpenNMS_and_the_Database
opennms[119134]: failed
systemd[1]: opennms.service: Control process exited, code=exited, status=1/FAILURE
systemd[1]: opennms.service: Failed with result 'exit-code'.
systemd[1]: Failed to start OpenNMS server.

You can also run config-tester -a to verify your configuration settings. If the tests pass, a success message is displayed.

Non-obvious error messages with older Linux distributions

Some older Linux distributions (Red Hat Enterprise Linux 7, CentOS 7, and so on) are more terse when a Meridian startup failure occurs. An additional diagnostic step can help reveal the problem.

After installing the software and running systemctl start opennms, the process immediately exits with a brief error message:

# systemctl start opennms
Job for opennms.service failed because the control process exited with error code.
See "systemctl status opennms.service" and "journalctl -xe" for details.

Running systemctl status opennms produces the following output:

systemd[1]: Starting OpenNMS server...
systemd[1]: opennms.service: control process exited, code=exited status=1
systemd[1]: Failed to start OpenNMS server.
systemd[1]: Unit opennms.service entered failed state.
systemd[1]: opennms.service failed.

You can also start the Meridian services verbosely; this provides additional details about any errors:

# $\{OPENNMS_HOME}/bin/opennms -v start
WARNING: relaunching as opennms
WARNING: `systemctl` was found, but OpenNMS is not being run in Systemd mode.
You will not be able to use systemctl to manage an OpenNMS started manually from the CLI.
Starting OpenNMS: /opt/opennms/bin/opennms: OpenNMS not configured.
/opt/opennms/etc/configured does not exist.

You need to run the installer to set up the database.  In most
cases, it is enough to run:

  /opt/opennms/bin/install -dis

For details, see the install guide at:

http://www.opennms.org/index.php/QuickStart#Initialize_OpenNMS_and_the_Database

[FAILED]

In the above example, ${OPENNMS_HOME}/bin/install -dis had not been run, causing startup to fail.

Database errors

If you encounter a database error when starting Meridian, database access may not have been configured properly. In this case, error messages are displayed when you run systemctl start opennms:

systemd[1]: Starting OpenNMS server...
opennms[133961]: Starting OpenNMS:
opennms[134703]: error invoking "check" operation: Data source 'opennms' failed.
opennms[133961]: OpenNMS was unable to connect to the 'opennms' database configured in opennms-datasources.xml.
opennms[133961]: failed
systemd[1]: opennms.service: Control process exited, code=exited, status=1/FAILURE
systemd[1]: opennms.service: Failed with result 'exit-code'.
systemd[1]: Failed to start OpenNMS server.

See Set up PostgreSQL for information on installing and setting up PostgreSQL.

rpm, yum, dnf upgrade errors

Sometimes, Meridian startup may fail because there are RHEL-related .rpmnew or .rpmsave files located in ${OPENNMS_HOME}. When you check the status of your instance, one of the following errors might be displayed:

.rpmnew error

opennms[8728]: WARNING!  You have files that end in .rpmnew in your
opennms[8728]: OPENNMS_HOME (/opt/opennms) directory.
opennms[8728]: The format of the original files may have changed since
opennms[8728]: you modified them before installing a new version.
opennms[8728]: Please double-check that your configuration files are
opennms[8728]: up-to-date and delete any leftover .rpmnew files or
opennms[8728]: OpenNMS will not start.
opennms[8011]: failed
systemd[1]: opennms.service: Control process exited, code=exited, status=1/FAILURE
systemd[1]: opennms.service: Failed with result 'exit-code'.
systemd[1]: Failed to start OpenNMS server.

.rpmsave error

opennms[133068]: WARNING!  You have files that end in .rpmsave in your
opennms[133068]: OPENNMS_HOME (/opt/opennms) directory.
opennms[133068]: The format of the original files may have changed since
opennms[133068]: you modified them before installing a new version.
opennms[133068]: Please double-check that your configuration files are
opennms[133068]: up-to-date and delete any leftover .rpmsave files or
opennms[133068]: OpenNMS will not start.
opennms[132345]: failed
systemd[1]: opennms.service: Control process exited, code=exited, status=1/FAILURE
systemd[1]: opennms.service: Failed with result 'exit-code'.
systemd[1]: Failed to start OpenNMS server.

You should evaluate the .rpmnew or .rpmsave file against the original file with the same name to determine which to keep. Delete the appropriate file, remove the .rpmnew or .rpmsave file extension (if applicable), and start Meridian.

You can run this command to identify any .rpmnew or .rpmsave files in your environment:

$ sudo find ${OPENNMS_HOME} -name '*.rpm*' -exec ls -la {} \;