I know the difference between 'active' and 'active and start PP'.
What I want to know is the relationship of this configuration at the sensor level, and at the method level.
These are what I mean by 'sensor level' and 'method level':
Is it like....one is of global level while another is of deeper level where you can choose to override the global level setting? If that's the case then which one is global setting?
Or, if I am wrong, what is the relationship between them?
Say, for this custom sensor, in the sensor rule (method level), I configure all method to be 'active and start PP', but in the sensor configuration of an agent group(sensor level), I configure it to be only 'active'. What is the result? The result would be active&start PP? or just active?
You are generally correct. The sensor configuration is for the entire sensor group (global to all the methods in the group) while the capture settings in the sensor group is specific to each method. You can activate and deactivate all the methods in a sensor group via the sensor configuration. You can also set it to have all the methods in the group to start PurePaths as well.
For cases where you want all the methods to be placed but only a few of the selections in the group to start PurePaths, you can change the capture settings to be as such and set the sensor configuration to be active much like you have pictured in your screenshots.
A similar thing applies with the placement checkbox next to the capture settings and the sensor placement check box in the agent group.
To reduce ambiguity and confusion, the best practice is to create a separate sensor group for the methods that start purepaths and to always also set "start purepaths" in the sensor configuration for that group too.
Name the sensor group SomethingMeaningful_EntryPoints so you know, and it reduces the confusion exponentially.
Also, I generally name my sensor groups based on the agent group they belong to, when the classes are unique to one agent group. This makes it easier to know which agent group to browse to add more classes/methods and to know exactly where to place the sensor group.
Thank you all for all the input! The reason I asked this is because, I was configuring custom sensor in a staging environment, where I can do a hot sensor placement whenever I want. Now that I've sucessfully create the useful custom sensor, I import it into the production environment, only to realize that I can't configure the sensor group to be 'active and start PP' (because I haven't do hot sensor placement yet, until the customer give me green light)
Hence, initially in the staging environment, here's how I configure:
Sensor Group - 'Active and start PP'
All the method within - 'active' only
Now, before I import it into production (and of course, before hot sensor placement/ app service restart):
Sensor Group - NA, since it doesn't appear in sensor placement config, yet
All the method within - change to 'active and start PP'
I hope the result would be the same. Afraid that it might not, thus, I asked this in the forum.
The outcome should be the same. I see that you configured it in the method to prevent having to change the sensor group after it is applied, that is an effective way of configuring the sensor.
Once you restart your app or hot sensor place the sensor, it will be added with the note to start PurePaths when that method is seen in AppMon.