I currently have an instance of 0.9.1.1 with Kafka=>Realtime ingestion.
Looking at the documentation (http://druid.io/docs/0.9.1.1/configuration/realtime.html) I’ve seen that I can also add Cache on Realtime nodes.
How will impact this cache? I’ve multiple datasources with million of rows ingested per hour (segment granularity is per hour).
Most of my queries interact with the Realtime nodes because belongs the status of current hour taking decision for next one.
Can the Realtime cache be a benefit with this use case?
Does it impact with Broker cache?
Which is the most appropriate cache for Realtime, local or memcache?
afaik realtime cache works only for local.
Thanks for the feedback. I can use the local cache.
But is the use case I’ve the one that benefits of this cache? How it interact with Broker cache system?
caching at realtime should suffice for your use case in my opinion.please note there is a parameter to control size of cache in real time.adjust it accordingly
Thanks for your feedback.
I’m implementing it.
Into the runtime.properties I’ve set these
and into common.runtime.properties