What does ‘very high load’ mean for you? –––– Sébastien Han Cloud Engineer
"Always give 100%. Unless you're giving blood.”
Phone: +33 (0)1 49 70 99 72
Mail: sebastien....@enovance.com
Address : 11 bis, rue Roquépine - 75008 Paris
Web : www.enovance.com - Twitter : @enovance
Begin forwarded message:
> From: "Ben's Gmail" <hengfengha...@gmail.com>
> Subject: Ceph perfomance issue!
> Date: 6 May 2014 11:25:17 GMT+2
> To: sebast...@enovance.com
>
> Hi,Sebastien!
>
> I recently work with Openstack(Havana),and Ceph(v0.72.2 Emperor). I used
> ceph to store all glance images,cinder volumes and vms.I boot vms from images
> create a new volume.
>
> There are 5 servers in my ceph cluster .4 servers run osds,each server has
> 5 disks, one for system(RHEL6.4),3 for osd deamon ,the other one is ssd for
> juernal .1 server run ceph moniter.All
> the server is connent with a bond divice (2X10G NIC).I install ceph with
> ceph-deploy. and I change the crush map.
> Here is my curshmap :
> # begin crush map
> tunable choose_local_tries 0
> tunable choose_local_fallback_tries 0
> tunable choose_total_tries 50
> tunable chooseleaf_descend_once 1
>
> # devices
> device 0 osd.0
> device 1 osd.1
> device 2 osd.2
> device 3 osd.3
> device 4 osd.4
> device 5 osd.5
> device 6 osd.6
> device 7 osd.7
> device 8 osd.8
> device 9 osd.9
> device 10 osd.10
> device 11 osd.11
>
> # types
> type 0 osd
> type 1 host
> type 2 rack
> type 3 row
> type 4 room
> type 5 datacenter
> type 6 root
>
> # buckets
> host F7-Cloud25 {
> id -2 # do not change unnecessarily
> # weight 1.650
> alg straw
> hash 0 # rjenkins1
> item osd.0 weight 0.550
> item osd.1 weight 0.550
> item osd.2 weight 0.550
> }
> host F7-Cloud26 {
> id -3 # do not change unnecessarily
> # weight 1.650
> alg straw
> hash 0 # rjenkins1
> item osd.3 weight 0.550
> item osd.4 weight 0.550
> item osd.5 weight 0.550
> }
> host F7-cloud27 {
> id -4 # do not change unnecessarily
> # weight 1.650
> alg straw
> hash 0 # rjenkins1
> item osd.7 weight 0.550
> item osd.8 weight 0.550
> item osd.6 weight 0.550
> }
> host F7-Cloud28 {
> id -5 # do not change unnecessarily
> # weight 1.650
> alg straw
> hash 0 # rjenkins1
> item osd.10 weight 0.550
> item osd.11 weight 0.550
> item osd.9 weight 0.550
> }
> root default {
> id -1 # do not change unnecessarily
> # weight 6.600
> alg straw
> hash 0 # rjenkins1
> item F7-Cloud25 weight 1.650
> item F7-Cloud26 weight 1.650
> item F7-cloud27 weight 1.650
> item F7-Cloud28 weight 1.650
> }
> rack A {
> id -6 # do not change unnecessarily
> # weight 3.300
> alg straw
> hash 0 # rjenkins1
> item F7-Cloud25 weight 1.650
> item F7-Cloud26 weight 1.650
> }
> host F7-Cloud27 {
> id -9 # do not change unnecessarily
> # weight 0.000
> alg straw
> hash 0 # rjenkins1
> }
> rack B {
> id -7 # do not change unnecessarily
> # weight 3.300
> alg straw
> hash 0 # rjenkins1
> item F7-cloud27 weight 1.650
> item F7-Cloud28 weight 1.650
> item F7-Cloud27 weight 0.000
> }
> root AB {
> id -8 # do not change unnecessarily
> # weight 6.600
> alg straw
> hash 0 # rjenkins1
> item A weight 3.300
> item B weight 3.300
> }
>
> # rules
> rule data {
> ruleset 0
> type replicated
> min_size 1
> max_size 10
> step take default
> step chooseleaf firstn 0 type host
> step emit
> }
> rule metadata {
> ruleset 1
> type replicated
> min_size 1
> max_size 10
> step take default
> step chooseleaf firstn 0 type host
> step emit
> }
> rule rbd {
> ruleset 2
> type replicated
> min_size 1
> max_size 10
> step take default
> step chooseleaf firstn 0 type host
> step emit
> }
> rule AB {
> ruleset 3
> type replicated
> min_size 1
> max_size 10
> step take AB
> step chooseleaf firstn 0 type rack
> step emit
> }
>
> # end crush map
>
>
>
> ##################################
>
>
>
> I create a ceph-test pool use eth crush rule 3,and test my ceph cluser
> with rados bench , I fount that only one osd host's load very high . I don't
> knew what to do .
>
>
>
> Thanks for your help!!!
signature.asc
Description: Message signed with OpenPGP using GPGMail
_______________________________________________ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
