Known Issues for Logging
These known issues have been identified in Logging.
Some agent warnings can be ignored
- Details
- Benign warnings might occur for the Oracle fluentd-based agent, similar to the following:
Sep 22 05:47:43 ociutv3mgftp02 ruby[1278962]: /opt/unified-monitoring-agent/embedded/lib/ruby/gems/2.6.0/gems/oci-2.9.0.1125/lib/oci/identity/models/base_tag_definition_validator.rb:23: warning: already initialized constant OCI::Identity::Models::BaseTagDefinitionValidator::VALIDATOR_TYPE_ENUM Sep 22 05:47:43 ociutv3mgftp02 ruby[1278962]: /opt/unified-monitoring-agent/embedded/lib/ruby/gems/2.6.0/gems/oci-2.9.0.1125/lib/oci/identity/models/base_tag_definition_validator.rb:24: warning: previous definition of VALIDATOR_TYPE_ENUM was here
- Workaround
- You can ignore benign warnings. These warnings have no impact on agent functionality.
Loading a new configuration in the Unified Monitoring Agent
- Details
-
Updates to a configuration might not be loaded because of a known issue with the configuration reload signal.
sudo journalctl -u unified-monitoring-agent --no-pager | grep -E "log writing failed|Failed to set up mount"
If running this command results in any of the following errors, a problem loading a new configuration in the Unified Monitoring Agent can be the cause.
fluentd[16817]: log writing failed. can't be called from trap context
or:
systemd[18174]: unified-monitoring-agent.service: Failed to set up mount namespacing: /run/systemd/unit-root/run/unified-monitoring-agent: No such file or directory
This problem doesn't break existing logging functionality, and the agent continues to run using the old configuration.
- Workaround
-
Restart the Unified Monitoring Agent.
sudo systemctl restart unified-monitoring-agent