From 3c54a49e39a823883b4f88cb1ebaa3cbd70d247f Mon Sep 17 00:00:00 2001 From: Dan Mick Date: Thu, 3 Apr 2014 11:24:33 -0700 Subject: [PATCH] Wordsmith the erasure-code doc a bit Signed-off-by: Dan Mick --- doc/dev/erasure-coded-pool.rst | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-) diff --git a/doc/dev/erasure-coded-pool.rst b/doc/dev/erasure-coded-pool.rst index 3dbd95698b333..8ae57d4321a57 100644 --- a/doc/dev/erasure-coded-pool.rst +++ b/doc/dev/erasure-coded-pool.rst @@ -4,10 +4,10 @@ Erasure Coded pool Purpose ------- -Erasure coded pools requires less storage space compared to replicated -pools. It has higher computation requirements and only supports a -subset of the operations allowed on an object (no partial write for -instance). +Erasure-coded pools require less storage space compared to replicated +pools. The erasure-coding support has higher computational requirements and +only supports a subset of the operations allowed on an object (for instance, +partial write is not supported). Use cases --------- @@ -15,7 +15,7 @@ Use cases Cold storage ~~~~~~~~~~~~ -An erasure coded pool is created to store a large number of 1GB +An erasure-coded pool is created to store a large number of 1GB objects (imaging, genomics, etc.) and 10% of them are read per month. New objects are added every day and the objects are not modified after being written. On average there is one write for 10,000 @@ -23,10 +23,10 @@ reads. A replicated pool is created and set as a cache tier for the replicated pool. An agent demotes objects (i.e. moves them from the -replicated pool to the erasure coded pool) if they have not been +replicated pool to the erasure-coded pool) if they have not been accessed in a week. -The erasure coded pool crush ruleset targets hardware designed for +The erasure-coded pool crush ruleset targets hardware designed for cold storage with high latency and slow access time. The replicated pool crush ruleset targets faster hardware to provide better response times. @@ -35,10 +35,10 @@ Cheap multidatacenter storage ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Ten datacenters are connected with dedicated network links. Each -datacenters contain the same amount of storage, have no power supply -backup and no air cooling system. +datacenter contains the same amount of storage with no power-supply +backup and no air-cooling system. -An erasure coded pool is created with a crush map ruleset that will +An erasure-coded pool is created with a crush map ruleset that will ensure no data loss if at most three datacenters fail simultaneously. The overhead is 50% with erasure code configured to split data in six (k=6) and create three coding chunks (m=3). With @@ -47,17 +47,17 @@ replication the overhead would be 400% (four replicas). Interface --------- -Set up an erasure coded pool:: +Set up an erasure-coded pool:: $ ceph osd create ecpool 12 12 erasure -Set up an erasure coded pool and the associated crush ruleset:: +Set up an erasure-coded pool and the associated crush ruleset:: $ ceph osd crush rule create-erasure ecruleset $ ceph osd pool create ecpool 12 12 erasure \ default ecruleset -Set the ruleset failure domain to osd instead of the host which is the default:: +Set the ruleset failure domain to osd (instead of the host which is the default):: $ ceph osd erasure-code-profile set myprofile \ ruleset-failure-domain=osd -- 2.39.5