On Fri, 18 Aug 2017 00:09:48 +0200 Mehmet wrote: > *resend... this Time to the list...* > Hey David Thank you for the response! > > My use case is actually only rbd for kvm Images where mostly Running Lamp systems on Ubuntu or centos. > All Images (rbds) are created with "proxmox" where the ceph defaults are used (actually Jewel in the near Future luminous...) > > What i want to know is Primary which constelation would be optimal for bluestore? > > I.e. > Put db and RAW Device on HDD and the Wal on nvme in my case? > Should i replace a HDD with an ssd and use this for the dbs so thats finally HDD as RAW Device, db on ssd and Wal on nvme? > Or... use the HDD for RAW and the nvme for Wal and db? > Ideally (based on what I've been reading here) you'll have your WALs on NVMe (with a 2nd NVMe to reduce the failure domain to 12 OSDs and improve IOPS/durability or RAID1 to remove a SPOF). Then you'll have the DB on decent (and larger) SSDs (not just one, same as above), as small writes will use this essentially as a journal again. And finally HDD OSDs as before. If you get a 2nd NVMe and use it individually for 12 OSDs, you could go for something like 24GB for the 12 WALs and 30GB each for the DBs, which would be a decent compromise I suppose. But the DB size feels a bit small, somebody with actual experience ought to pipe up here. Christian > Hope you (and others) understand what i mean :) > > - Mehmet > > Am 16. August 2017 19:01:30 MESZ schrieb David Turner <drakonstein@xxxxxxxxx>: > >Honestly there isn't enough information about your use case. RBD usage > >with small IO vs ObjectStore with large files vs ObjectStore with small > >files vs any number of things. The answer to your question might be > >that > >for your needs you should look at having a completely different > >hardware > >configuration than what you're running. There is no correct way to > >configure your cluster based on what hardware you have. What hardware > >you > >use and what configuration settings you use should be based on your > >needs > >and use case. > > > >On Wed, Aug 16, 2017 at 12:13 PM Mehmet <ceph@xxxxxxxxxx> wrote: > > > >> :( no suggestions or recommendations on this? > >> > >> Am 14. August 2017 16:50:15 MESZ schrieb Mehmet <ceph@xxxxxxxxxx>: > >> > >>> Hi friends, > >>> > >>> my actual hardware setup per OSD-node is as follow: > >>> > >>> # 3 OSD-Nodes with > >>> - 2x Intel(R) Xeon(R) CPU E5-2603 v3 @ 1.60GHz ==> 12 Cores, no > >>> Hyper-Threading > >>> - 64GB RAM > >>> - 12x 4TB HGST 7K4000 SAS2 (6GB/s) Disks as OSDs > >>> - 1x INTEL SSDPEDMD400G4 (Intel DC P3700 NVMe) as Journaling Device > >for > >>> 12 Disks (20G Journal size) > >>> - 1x Samsung SSD 840/850 Pro only for the OS > >>> > >>> # and 1x OSD Node with > >>> - 1x Intel(R) Xeon(R) CPU E5-2650 v3 @ 2.30GHz (10 Cores 20 Threads) > >>> - 64GB RAM > >>> - 23x 2TB TOSHIBA MK2001TRKB SAS2 (6GB/s) Disks as OSDs > >>> - 1x SEAGATE ST32000445SS SAS2 (6GB/s) Disk as OSDs > >>> - 1x INTEL SSDPEDMD400G4 (Intel DC P3700 NVMe) as Journaling Device > >for > >>> 24 Disks (15G Journal size) > >>> - 1x Samsung SSD 850 Pro only for the OS > >>> > >>> As you can see, i am using 1 (one) NVMe (Intel DC P3700 NVMe – 400G) > >>> Device for whole Spinning Disks (partitioned) on each OSD-node. > >>> > >>> When „Luminous“ is available (as next LTE) i plan to switch vom > >>> „filestore“ to „bluestore“ 😊 > >>> > >>> As far as i have read bluestore consists of > >>> - „the device“ > >>> - „block-DB“: device that store RocksDB metadata > >>> - „block-WAL“: device that stores RocksDB „write-ahead journal“ > >>> > >>> Which setup would be usefull in my case? > >>> I Would setup the disks via "ceph-deploy". > >>> > >>> Thanks in advance for your suggestions! > >>> - Mehmet > >>> ------------------------------ > >>> > >>> 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 > >> -- Christian Balzer Network/Systems Engineer chibi@xxxxxxx Rakuten Communications _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |