Max_snapshot_bytes_per_sec - Implications of setting this value

I am after a more in depth explanation of the implications of adjusting the max_snapshot_bytes_per_sec for an S3 repository. Reading the documentation, it is simply described as

(Optional, byte value) Maximum snapshot creation rate per node. Defaults to 40mb per second.

My main questions are:

  • Does adjusting this setting have a material impact on the cluster? ie: If I set this to way too high, is the Snapshotting process going to consume more resources and start to block index and search operations? I have found forum posts on here and stack overflow recommending increasing it (some increasing it to 10G).
  • Does changing this setting impact running snapshots, or only new ones?

I have a cluster with 4 data nodes (to expand to probably 6) and around 20 TB of data (to expand to ~40) and want to ensure I am getting sufficient snapshot throughput to actually have a functional backup. I forgot to adjust my snapshot frequency as the data volumes have increased, and as a result I have spent 2 days just trying to get snapshots completing again and would like to know if adjusting this setting is a recommended approach to improving throughput.


This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.