Making Zstd useful for more clusters
21 Mar
2020
21 Mar
'20
3:06 a.m.
A few months ago I created a feature request to lower the default Zstd compression level to 1 and make it configurable. This fixes the performance issue that prevents us from using Zstd for compression on our clusters (the current default in Ceph is compression level 5, but the Zstd upstream uses 3): https://tracker.ceph.com/issues/43377 There weren't any takers, so I created my own fix that I would like to see merged and backported to Nautilus and Octopus: https://github.com/ceph/ceph/pull/33790 Let me know what you guys think and if there are any changes you'd like to see. Thanks, Bryan
2383
Age (days ago)
2383
Last active (days ago)
0 comments
1 participants
participants (1)
-
Bryan Stillwell