DX NetOps Performance Management Data Collector (DC) doesn't run.
Data Collector seems to be running but according to System Health Dashboards the DC is flatlined and is not polling any data.
Data Collector not showing new polled data.
Data Collector fails to stay running after being restarted.
Data Collector dcmd service fails to stay running.
In the DC's (default path) /opt/IMDataCollector/apache-karaf-<version>/data/log/karaf.log file the following errors will be present.
Older releases may show:
java.io.IOException: Frame size of 169 MB larger than max allowed 100 MB
at org.apache.activemq.openwire.OpenWireFormat.unmarshal(OpenWireFormat.java:269)[162:org.apache.activemq.activemq-osgi:5.13.1]
at org.apache.activemq.transport.tcp.TcpTransport.readCommand(TcpTransport.java:240)[162:org.apache.activemq.activemq-osgi:5.13.1]
at org.apache.activemq.transport.tcp.TcpTransport.doRun(TcpTransport.java:232)[162:org.apache.activemq.activemq-osgi:5.13.1]
at org.apache.activemq.transport.tcp.TcpTransport.run(TcpTransport.java:215)[162:org.apache.activemq.activemq-osgi:5.13.1]
at java.lang.Thread.run(Thread.java:745)[:1.8.0_72]
<DATE><TIME> | INFO | ActiveMQ Task-2 | FailoverTransport | sport.failover.FailoverTransport 1077 | 162 - org.apache.activemq.activemq-osgi - 5.13.1 | | Successfully reconnected to tcp://127.0.0.1:61616
Newer releases may show:
2022-02-23 15:34:40,863 | INFO | ActiveMQ Task-2 | FailoverTransport | sport.failover.FailoverTransport 1054 | 165 - org.apache.activemq.activemq-osgi - 5.15.8 | | Successfully reconnected to tcp://127.0.0.1:61616
2022-02-23 15:34:40,866 | WARN | .0.1:61616@55678 | FailoverTransport | sport.failover.FailoverTransport 280 | 165 - org.apache.activemq.activemq-osgi - 5.15.8 | | Transport (tcp://127.0.0.1:61616) failed , attempting to automatically reconnect: java.io.IOException: Frame size of 137 MB larger than max allowed 100 MB
All supported DX NetOps Performance Management Data Collector releases
To resolve this we'll need to increase the maxFrameSize configuration to a value higher than the highest value reported by the log messages.
The dcmd DC service should now remain running.
Is there any downside to increasing the maxFrameSize?
When increasing maxFramSize speculation suggests that it would use more memory for a messages in AMQ, and possibly take a little (micro or milli seconds) longer to send the message to the DA. The collection of java memory garbage collection may come into play, as the DC will need a bigger block of memory to store the messages, and will then garbage collect it later. Best not to go bigger than 250 MB in total.