Must-Gather scripts are required for the sensors

The Tiger Team is working with support an others on a tool called Tiger-Optics that is using AI to  analyze Must-Gather data from customers to help accelerate problem diagnostics.   So far we have done work to process data for the backend.  This is partly due to the fact that we have a must-gather tool to collect data from kubernetes and the Instana backend.

What is lacking today is a consistent way of gathering Sensor Must-Gather data from a customer.   Many of the sensors have documented must-gather steps here:  https://www.ibm.com/support/pages/mustgather-read-first-Instana  

Today, we have a partial mechanism to collect the Agent must gather and an option to collect the logs.   There is a 2nd Idea to combine all of the relevant data into a single script.   

Similar scripts are necessary for the sensors.  What we would like to see is a structure that is architected in a way that makes it intuitive for support when analyzing a support case and also allows our AI solutions to parse through the config files, logs, etc.

The structure might look something like this...keeping in mind that there will be multiple sensors on a machine:

•Sensor Name (example MQ)

•Middleware logs

•middleware config files

•Product version/environment

•In the case of MQ, there are additional files like the User Exit that gets installed

•There are also possible commands and command outputs to gather additional information…but I suspect they won’t be able to automate them.

•ACE sensor

•…

•Java sensor

•…

•etc.

Some data will likely need to be gathered as root.  Other data is likely to be gathered as the middleware owner.   Ideally, we would like a single script that would prompt for login credential or sudo permissions.  But, if necessary, multiple scripts are okay as long as they build the proper directory structure and are consistent.