So is there any way, through some commands, that I can see how many objects go through the WAL & DB devices, and how optimal is performs?
On Tue, Nov 21, 2017 at 10:52 AM, Rudi Ahlers <[email protected]> wrote: > Hi, > > I didn't see where you said there is no journal on bluestore. > > Maybe I'm confusing the terminology. I have created a DB and WAL device > for my Bluestore, but I presume that's not a journal (anymore?). > > This is how I set it up this time: ceph-disk prepare --bluestore /dev/sdc > --block.wal /dev/sde --block.db /dev/sde > > Apart from "ceph grafana", which I can't install at this stage since the > cluster isn't ready yet, how else can I test the SSD cache drives? > > On Tue, Nov 21, 2017 at 10:37 AM, Christian Balzer <[email protected]> wrote: > >> On Tue, 21 Nov 2017 10:23:20 +0200 Rudi Ahlers wrote: >> >> > Hi, >> > >> > Is it possible to test the CEPH journal to see if it performs optimally? >> > >> You're using bluestore, as I said before, there is no journal. >> >> The SSDs you're using are top of the line, perfectly suited for what >> you're using them for. >> >> > And, how can I see how much data is being cached? >> > >> Nothing gets cached per se with bluestore. >> As for utilization, there were values for the journal, AFAIK there isn't >> anything for the DB stuff at this point. >> >> In general, google "ceph grafana", you will REALLY want to monitor a ceph >> cluster this way. >> >> -- >> Christian Balzer Network/Systems Engineer >> [email protected] Rakuten Communications >> > > > > -- > Kind Regards > Rudi Ahlers > Website: http://www.rudiahlers.co.za > -- Kind Regards Rudi Ahlers Website: http://www.rudiahlers.co.za
_______________________________________________ ceph-users mailing list [email protected] http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
