Ok, so now my app has a scripted input (Powershell) and setup.xml is communicating with an eai endpoint defined inside the app. That's great, but the question is how should I manage the configuration of my scripted input such as target's hostname, user credentials, etc?
My first thought was to keep this information inside a plain xml file and have my script and the endpoint access it. But I guess this solution will make the settings global and won't work scale for multiple users, so I think I'm expected to use *.conf files instead. Plus it adds some inconsistency to the way Splunk keeps its configs.
Going this way to collect the actual config in my script I should not only read a conf file from default\, but also merge it with a conf from local\ and, possibly, respect a conf inside users\someuser\myapp\my.conf.
The last part seems confusing to me. How can I tell from my script which user's .conf should I read?
Anyway that's just too many of details to implement in my simple script. I believe there should be a simpler way to manage config without duplicating what Splunk already can do. I guess my script could read the config from my app's EAI endpoint, but for this it needs to authenticate first. If I could pass the script a session key somehow that would be great..
What would you recommend? Just stick with a simple xml?
thanks 🙂
