Performance impact during snapshot process


We have a cluster of ES indices in which we are storing logstash logs ( with one index per day ). Each index has multiple replicas.

  1. Is there any performance impact on the querying and indexing of the current logstash indices when snapshot & restore going in parallel for some old indices (in which no read or write is happening let's say)?
  2. What is the performance impact snapshot is taken from an index in which querying and indexing as happening simultaneously?

(Mark Walkom) #2

Yes there is increased disk activity, for example, as data is read off. This is something you need to test on your own data to find out what the impact is.


Hmm. Thanks @warkolm

(system) #4