AS400 as I now is native technology and should be instrumented manually.
In Dynatrace you can create Custom Network Device which will be representation of AS400 in dynatrace. Then you can push metrics from AS400 over API (if it is possible somehow) to this custom network device. This device will be visible as process and service with proper connections in your infrastructure. But this will not give you code level visibility.
Unfortunately I can't provide you any.
Here is info about using custom network device:
You can also push there metrics using API. In such case you will no longer see in Dynatrace requests to unmonitored hosts but you will see requests to AS400 (if this will be name of your custom network device). But you will not see any transactions on AS400.
If AS400 was created as a "Custom Network Device",
how would it be associated to the database or application?
Would Dynatrace target the AS400
server as the root cause of a degradation problem? (high CPU usage, low free memory)
I think it would be easier to recover the metrics of some
monitoring platform with API support (e.g. Orion)
We're currently in an EAP stage of the iSeries (as400) monitoring. You can read more about it here:
Someone can share more technical information than the one published in dynatrace documentation, about what difference it has to configure QPFRDATA versus QMPGDATA, or select the check "run Performance collection Service ...."
When the customer already has a monitoring data collection process running, how can I activate our plugin without the customer losing its monitoring data?
Thanks a lot
The difference between QPFRDATA and QMPGDATA is really just the storage locations. In some cases, depending on the configuration of the AS400, collection libraries are stored in QPFRDATA and others QMPGDATA. The drop-down gives you the choice to select which collection library the plugin must look into. It all depends how the AS400 was configured.
The "Run Performance Collection service" checkbox sends a command to the AS400 machine (CFGPFRCOL command) and then sends the STRPFRCOL command. The CFGPFRCOL command will overwrite the existing (if any) custom performance collection configuration. That is why it is optional as a checkbox.
If the customer already has a monitoring collection process running, for example they have another tool already collecting data, then it is best to not check that checkbox. The configuration that the plugin uses aggregates data per 1 minute intervals on the AS400, because the plugin polls 1 minute data. So if the existing monitoring performance collection is aggregating data for 10 minute intervals, then the plugin would overwrite that configuration and may interfere with the other monitoring tool. However, if you don't use Performance Collection Services, then you will get less metrics back because the plugin would default to using metrics from the AS400 catalog tables which does not have Disk performance metrics for example.