<br><br><div class="gmail_quote">On Wed, Sep 12, 2012 at 12:05 AM, siga hiro <span dir="ltr"><<a href="mailto:hirokisiga@gmail.com" target="_blank">hirokisiga@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Hi.<br>
<br>
I make the data area of postgresql in glusterfs on the server.<br>
And,I mounted a glusterfs area from postgres-node, and postgresql started.<br>
Then,It became an error when pgbench start.<br>
( if fuse-mount,became error.if nfs-mount,not error.<br>
mount -t glusterfs 172.22.0.181:/syncdata /syncdata -> pgbench ERROR<br>
mount -t nfs -o nolock,vers=3,tcp,soft,timeo=<a href="tel:3%20172.22.0.181" value="+13172220181">3 172.22.0.181</a>:/syncdata<br>
/syncdata -> pgbench OK<br>
)<br>
<br>
I found this [<a href="https://bugzilla.redhat.com/show_bug.cgi?id=810944]." target="_blank">https://bugzilla.redhat.com/show_bug.cgi?id=810944].</a>..<br>
<br>
Is it the bug of the kernel, or can i cope with it with glusterfs, or<br>
is there what or an idea would you teach?<br>
<br>
<br>
Node:<br>
postgres node:172.22.0.180<br>
OS:ScientificLinux61<br>
DB:postgresql91-9.1.4-3PGDG<br>
<br>
glusterfs node <a href="http://172.22.0.181/172.22.0.182" target="_blank">172.22.0.181/172.22.0.182</a><br>
OS:ScientificLinux61<br>
glusterfs3.3.0-1<br>
<br>
Details:<br>
<br>
・I make the data area of postgresql on glusterfs.<br>
# gluster peer status -> 172.22.0.181<br>
Number of Peers: 1<br>
Hostname: 172.22.0.182<br>
Uuid: 601a3cf7-8614-4b89-83e4-2dab319f3582<br>
State: Peer in Cluster (Connected)<br>
<br>
# gluster volume info<br>
Volume Name: syncdata<br>
Type: Replicate<br>
Volume ID: c7708db1-0eb7-4a0d-a21c-24162ac3ed8f<br>
Status: Started<br>
Number of Bricks: 1 x 2 = 2<br>
Transport-type: tcp<br>
Bricks:<br>
Brick1: 172.22.0.181:/home/syncdata<br>
Brick2: 172.22.0.182:/home/syncdata<br>
<br>
postgresql node<br>
# mount -t glusterfs 172.22.0.181:/syncdata /syncdata<br>
# ls /syncdata/pgdata/<br>
PG_VERSION global pg_ident.conf pg_serial pg_tblspc<br>
postgresql.conf<br>
base pg_clog pg_multixact pg_stat_tmp pg_twophase<br>
postmaster.opts<br>
pg_hba.conf pg_notify pg_subtrans pg_xlog postmaster.pid<br>
<br>
・Then, pgbench was started.<br>
postgresql node<br>
$ /usr/pgsql-9.1/bin/pgbench -i -s 100<br>
.....<br>
2012-09-12 11:53:51 ERROR: unexpected data beyond EOF in block<br>
99480 of relation base/12780/16396<br>
<br>
$ /usr/pgsql-9.1/bin/pgbench -h 172.22.0.182 -c 100 -t 10<br>
2012-09-12 11:53:51 ERROR: unexpected data beyond EOF in block<br>
99480 of relation base/12780/16396</blockquote><div><br></div><div>Can you check if any of the following options fix your problem -</div><div><br></div><div>1) --direct-io-mode=enable (to glusterfs)</div><div>2) --attribute-timeout=0 (to glusterfs)</div>
<div>3) gluster volume set syncdata performance.write-behind off</div><div><br></div><div>Thanks,</div><div>Avati</div><div> </div></div>