Hi everybody,
I underdstand that radosgw bucket indexes are stored as leveldb files.
Do writes leveldb go though the jounrnal first, or directly to disk?
In other words, do they get the benefits of having a fast journal?
I started investigating disk because I observe a lot of writes to ldb files
Hi,
we have a Ceph cluster with:
- 12 OSDs on 6 physical nodes, 64 GB RAM
- each OSD has a 6 TB spinning disk and a 10GB journal in ram (tmpfs) [1]
- 3 redundant copies
- 25% space usage so far
- ceph 0.94.2.
- store data via radosgw, using sharded bucket indexes (64 shards).
- 500 PGs per node (a