Hi, Another consideration probably is the size of the packet sent over the network. Samuel Just chose this default value but I never asked about the rationale. Regarding the stripe width, it actually is stored per pool although it does not look that way. If you want to configure a pool with a different stripe width you can ceph osd pool create poolA 12 12 erasure ceph tell mon.* injectargs "--osd-pool-erasure-code-stripe-width $((16 * 1024))" ceph osd pool create poolB 12 12 erasure ceph osd dump | grep '^pool ' ... pool 1 'poolA' erasure size 3 min_size 2 crush_ruleset 1 object_hash rjenkins pg_num 12 pgp_num 12 last_change 25 flags hashpspool stripe_width 4096 pool 2 'poolB' erasure size 3 min_size 2 crush_ruleset 1 object_hash rjenkins pg_num 12 pgp_num 12 last_change 27 flags hashpspool stripe_width 16384 that is : change the default value before creating the pool. The stripe size is stored as a property of the pool and cannot be changed after it is created. Cheers On 08/07/2014 07:23, Ma, Jianpeng wrote: > Hi Loic, > At present, I test different stripe-width(4k,16k,32k,256k,512k, 1024k) with k=3/m=2. I used cosbench+librados to test the performance with 10MB-write(object size). > > Stripe-width avg-cpu-uitl-of-osd(user+kernel) Throghput > 4k 83% 90.01 > 16k 79% 105.01 > 32k 76% 106.7 > 64k 72% 107.69 > 256k 68% 106.52 > 512k 68% 103.12 > 1024K 69% 102.57 > > >>From the result, we can know 4k not the best. > I know if the stripe-width increase, we should pad more data. Is there other considerations? > At present, we can't set different stripe-width for different erasure-code pool. Why? > > > Thanks! > Jianpeng Ma > -- Lo?c Dachary, Artisan Logiciel Libre -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 263 bytes Desc: OpenPGP digital signature URL: <http://lists.ceph.com/pipermail/ceph-users-ceph.com/attachments/20140708/47e27f63/attachment.pgp>
 |