We have two search heads:
- First is used with Enterprise Security with CIM installed and acceleration enabled on some datamodels
- On the second one(no enterprise security) we also have installed CIM, but manually, and haven't enabled acceleration on any datamodel
What would be the impact of having different acceleration settings enabled for each search head?
In our environment we run a 5 node cluster as one search head, a few single instance search heads and one ES, one ITSI heads. Acceleration is on ES and ITSI but not the others. There is no impact to the other search heads, but the accelerated data is only available on the head it is accelerated from,
I am a SplunkCloud customer, so they are very involved with the 'architecture' of our environment. Having the data accelerated where it is to be used aligns with their best practices.