DistributedMapCacheServer usage at ProcessGroup levels

[#1]

I have a data flow confined to a single ProcessGroup and am currently using DistributedMapCacheServer at the ProcessGroup level.

However, moving forward I need move the above process group as a child 1 ProcessGroup inside another ProcessGroup alongside another ProcessGroup which will be child 2 ProcessGrooup. So essentially, a parent ProcessGroup having 2 child ProcessGroups (or data flows) inside it. Apart from this, I need to configure another DistributedMapCacheServer at the Parent ProcessGroup level, which can be leveraged by both the child ProcessGroups, respectively.

So, my question is, since there are in total 2 DistributedMapCacheServer which will be used - one at the Parent ProcessGroup level and another at child 1 ProcessGroup level, is it OK to both maintain consistency and keep the Memory (planning to bump from current 4 GB to 8 GB instance) aspect of it? Is there an example Flow file for the same or someone tried that at POC or PROD level before?

Please advice. Thank you.

[#2 · Accepted solution]

@Avish Saha

Why the need for multiple Distributed Map Cache Servers? The server runs independent of the process group it was created under. It is bound to a port and is available for connections from any properly configured Distributed Map Cache client service. You can also configure one Distributes Map Cache Client Service at the parent process group level and it will be available to any sub process group, but there is nothing wrong with creating multiple clients either.

Thanks,

Matt