Maybe a bit of topic, just curious what speeds did you get previously? Depending on how you test your native drive of 5400rpm, the performance could be similar. 4k random read of my 7200rpm/5400 rpm results in ~60iops at 260kB/s. I also wonder why filestore could be that much faster, is this not something else? Maybe some dangerous caching method was on? -----Original Message----- From: Stuart Longland [mailto:stuartl@xxxxxxxxxxxxxxxxxx] Sent: vrijdag 19 juli 2019 12:22 To: ceph-users Subject: Future of Filestore? Hi all, Earlier this year, I did a migration from Ceph 10 to 12. Previously, I was happily running Ceph v10 on Filestore with BTRFS, and getting reasonable performance. Moving to Ceph v12 necessitated a migration away from this set-up, and reading the documentation, Bluestore seemed to be "the way", so a hasty migration was performed and now my then 3-node cluster moved to Bluestore. I've since added two new nodes to that cluster and replaced the disks in all systems, so I have 5 WD20SPZX-00Us storing my data. I'm now getting about 5MB/sec I/O speeds in my VMs. I'm contemplating whether I migrate back to using Filestore (on XFS this time, since BTRFS appears to be a rude word despite Ceph v10 docs suggesting it as a good option), but I'm not sure what the road map is for supporting Filestore long-term. Is Filestore likely to have long term support for the next few years or should I persevere with tuning Bluestore to get something that won't be outperformed by an early 90s PIO mode 0 IDE HDD? -- Stuart Longland (aka Redhatter, VK4MSL) I haven't lost my mind... ...it's backed up on a tape somewhere. _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |