Le 13/08/2018 à 16:29, Jason Dillaman a écrit : > > For such a small benchmark (2 GiB), I wouldn't be surprised if you are > not just seeing the Filestore-backed OSDs hitting the page cache for > the reads whereas the Bluestore-backed OSDs need to actually hit the > disk. Are the two clusters similar in terms of the numbers of > HDD-backed OSDs? >
New cluster has a bit more OSDs, and better hardware (raid card with cache, more memory, more cpu, and less workload). Old: # ceph osd tree ID WEIGHT TYPE NAME UP/DOWN REWEIGHT PRIMARY-AFFINITY -1 17.28993 root default -2 3.63998 host hyp-prs-01 0 1.81999 osd.0 up 1.00000 1.00000 1 1.81999 osd.1 up 1.00000 1.00000 -3 3.63997 host hyp-prs-02 3 1.81999 osd.3 up 1.00000 1.00000 2 1.81998 osd.2 up 1.00000 1.00000 -4 4.54999 host hyp-prs-03 4 1.81999 osd.4 up 1.00000 1.00000 5 2.73000 osd.5 up 1.00000 1.00000 -5 5.45999 host hyp-prs-04 6 2.73000 osd.6 up 1.00000 1.00000 7 2.73000 osd.7 up 1.00000 1.00000 New: # ceph osd tree ID CLASS WEIGHT TYPE NAME STATUS REWEIGHT PRI-AFF -1 43.66919 root default -3 14.55640 host osd-01 0 hdd 3.63910 osd.0 up 1.00000 1.00000 1 hdd 3.63910 osd.1 up 1.00000 1.00000 2 hdd 3.63910 osd.2 up 1.00000 1.00000 3 hdd 3.63910 osd.3 up 1.00000 1.00000 -5 14.55640 host osd-02 4 hdd 3.63910 osd.4 up 1.00000 1.00000 5 hdd 3.63910 osd.5 up 1.00000 1.00000 6 hdd 3.63910 osd.6 up 1.00000 1.00000 7 hdd 3.63910 osd.7 up 1.00000 1.00000 -7 14.55640 host osd-03 8 hdd 3.63910 osd.8 up 1.00000 1.00000 9 hdd 3.63910 osd.9 up 1.00000 1.00000 10 hdd 3.63910 osd.10 up 1.00000 1.00000 11 hdd 3.63910 osd.11 up 1.00000 1.00000 do you mean that with bluestore there is no page cache involved?
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
