<div dir="ltr"><div>Could you disable cluster.eager-lock and try again?<br><br></div>-Krutika<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Sep 21, 2017 at 6:31 PM, Lindsay Mathieson <span dir="ltr">&lt;<a href="mailto:lindsay.mathieson@gmail.com" target="_blank">lindsay.mathieson@gmail.com</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Upgraded recently from 3.8.15 to 3.10.5 and have seen a fairly substantial drop in read/write perfomance<br>
<br>
env:<br>
<br>
- 3 node, replica 3 cluster<br>
<br>
- Private dedicated Network: 1Gx3, bond: balance-alb<br>
<br>
- was able to down the volume for the upgrade and reboot each node<br>
<br>
- Usage: VM Hosting (qemu)<br>
<br>
- Sharded Volume<br>
<br>
- sequential read performance in VM&#39;s has dropped from 700Mbps to 300mbs<br>
<br>
- Seq Write has dropped from 115MB/s (approx) to 110<br>
<br>
- Write IOPS have dropped from 12MB/s to 8MB/s<br>
<br>
Apart from increasing the op version I made no changes to the volume settings.<br>
<br>
op.version is 31004<br>
<br>
gluster v info<br>
<br>
Volume Name: datastore4<br>
Type: Replicate<br>
Volume ID: 0ba131ef-311d-4bb1-be46-596e83<wbr>b2f6ce<br>
Status: Started<br>
Snapshot Count: 0<br>
Number of Bricks: 1 x 3 = 3<br>
Transport-type: tcp<br>
Bricks:<br>
Brick1: vnb.proxmox.softlog:/tank/vmda<wbr>ta/datastore4<br>
Brick2: vng.proxmox.softlog:/tank/vmda<wbr>ta/datastore4<br>
Brick3: vnh.proxmox.softlog:/tank/vmda<wbr>ta/datastore4<br>
Options Reconfigured:<br>
transport.address-family: inet<br>
cluster.locking-scheme: granular<br>
cluster.granular-entry-heal: yes<br>
features.shard-block-size: 64MB<br>
network.remote-dio: enable<br>
cluster.eager-lock: enable<br>
performance.io-cache: off<br>
performance.read-ahead: off<br>
performance.quick-read: off<br>
performance.stat-prefetch: on<br>
performance.strict-write-order<wbr>ing: off<br>
nfs.enable-ino32: off<br>
nfs.addr-namelookup: off<br>
nfs.disable: on<br>
cluster.server-quorum-type: server<br>
cluster.quorum-type: auto<br>
features.shard: on<br>
cluster.data-self-heal: on<br>
performance.readdir-ahead: on<br>
performance.low-prio-threads: 32<br>
user.cifs: off<br>
performance.flush-behind: on<br>
server.event-threads: 4<br>
client.event-threads: 4<br>
server.allow-insecure: on<span class="HOEnZb"><font color="#888888"><br>
<br>
<br>
-- <br>
Lindsay Mathieson<br>
<br>
______________________________<wbr>_________________<br>
Gluster-users mailing list<br>
<a href="mailto:Gluster-users@gluster.org" target="_blank">Gluster-users@gluster.org</a><br>
<a href="http://lists.gluster.org/mailman/listinfo/gluster-users" rel="noreferrer" target="_blank">http://lists.gluster.org/mailm<wbr>an/listinfo/gluster-users</a><br>
</font></span></blockquote></div><br></div>