[Gluster-users] High I/O And Processor Utilization

Lindsay Mathieson lindsay.mathieson at gmail.com
Sat Jan 9 23:11:04 UTC 2016


On 10/01/2016 1:44 AM, Kyle Harris wrote:
> I can make the change to sharding and then export/import the VMs to 
> give it a try.  So just to be clear, I am using v3.7.6-1.  Is that 
> sufficient?  I would rather not have to compile from source and would 
> probably wait for the next rpms if that is needed.


Speaking not as a dev (I'm not), but as a tester/user, 3.7.6 will do for 
testing & usage, thats what I am on. Write performance could be better 
and I believe there are some fixes for that due in 3.7.7


>
> Also, given the output below. what would you recommend I use for the 
> shard block size and furthermore, how do you determine this?

features.shard: on
features.shard-block-size: <size>

Where size takes std unit sizes, e.g 64M, 1G etc. Default is 4M. Shard 
size also has some interesting implications for the upcoming SSD Tier 
volumes (3.8). They are available in 3.7, if you are ok with regular 
breakages :)

I bench tested a lot and didn't find all that much difference in results 
for shard sizes. This is my current settings:

Volume Name: datastore1
Type: Replicate
Volume ID: 1261175d-64e1-48b1-9158-c32802cc09f0
Status: Started
Number of Bricks: 1 x 3 = 3
Transport-type: tcp
Bricks:
Brick1: vnb.proxmox.softlog:/vmdata/datastore1
Brick2: vng.proxmox.softlog:/vmdata/datastore1
Brick3: vna.proxmox.softlog:/vmdata/datastore1
Options Reconfigured:
features.shard: on
features.shard-block-size: 64MB
cluster.self-heal-window-size: 256
server.event-threads: 4
client.event-threads: 4
cluster.quorum-type: auto
cluster.server-quorum-type: server
performance.io-thread-count: 32
performance.cache-refresh-timeout: 4
nfs.disable: on
nfs.addr-namelookup: off
nfs.enable-ino32: off
performance.write-behind: on
performance.strict-write-ordering: on
performance.stat-prefetch: off
performance.quick-read: off
performance.read-ahead: off
performance.io-cache: off
cluster.eager-lock: enable
network.remote-dio: enable
performance.readdir-ahead: on
performance.write-behind-window-size: 256MB
performance.cache-size: 256MB


-- 
Lindsay Mathieson



More information about the Gluster-users mailing list