<div dir="ltr">Thank you guys for the response. <div><br></div><div><font color="#000000">I've configured 2(replica) x 2 (distributed) gluster node now. The reason why I tested with 1 node is that I wanted to check primitive performance for 1 node. We wanna make subscription with RedHat but Red branch here is not supporting GlusterFS yet. So I'm asking you guys... </font></div>
<div><font color="#000000"><br></font><div>Anyway..</div><div><span style="color:rgb(0,0,0)">I've check it larger block size using dd and iozone. (I used 4096 before)</span><br></div><div><font color="#000000">When it comes to 8192 byte or bigger as block size(record), the performance get much better like 80MB/s. </font></div>
<div><font color="#000000"><br></font></div><div><span style="color:rgb(0,0,0)"><br></span></div><div><span style="color:rgb(0,0,0)">So now I'm looking for<b> how to increase block size when JAVA app writes on file</b>. </span><span style="color:rgb(0,0,0)"> I've tried to increase write buffer size but it didn't make any change.</span></div>
<div><span style="color:rgb(0,0,0)"><br></span></div><div><span style="color:rgb(0,0,0)">I wonder if it's possible to adjust write block size for Java application. (like I did with dd tool adjusting block size..)</span></div>
<div><span style="color:rgb(0,0,0)">Is there any option on java platform? or Is there any way to adjust write block size for java app? </span></div><div><span style="color:rgb(0,0,0)"><br></span></div><div><span style="color:rgb(0,0,0)">If you have any suggestion, it will be very thankful.</span></div>
<div><span style="color:rgb(0,0,0)"><br></span></div><div>Regards,</div><div>Youngseok Jung</div><div><font color="#000000"><br>
</font><div class="gmail_extra"><font color="#000000"><br></font><br><div class="gmail_quote">2013/10/26 Anand Avati <span dir="ltr"><<a href="mailto:avati@gluster.org" target="_blank">avati@gluster.org</a>></span><br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div dir="ltr">Also, have you specified a block size for dd? The default (512 bytes) is too low for the number of context switches it generates in FUSE. Use a higher block size (64-128KB) and check the throughput.<div>
<br>
</div><div>Avati</div></div><div class="gmail_extra"><br><br><div class="gmail_quote"><div><div>On Fri, Oct 25, 2013 at 7:53 AM, Joe Julian <span dir="ltr"><<a href="mailto:joe@julianfamily.org" target="_blank">joe@julianfamily.org</a>></span> wrote:<br>
</div></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div><div>Have you brought this up with Red Hat Support? That is what you pay them for.<br>
<div><br>
<br>
<br>
Jung Young Seok <<a href="mailto:jung.youngseok@gmail.com" target="_blank">jung.youngseok@gmail.com</a>> wrote:<br>
>I've wrote below email. However it seems I missed mail key word rule on<br>
>subject.<br>
>So I'm sending it again.<br>
>Please check the below mail and any response will be helpful.<br>
>Thanks,<br>
>2013. 10. 25. ¿ÀÈÄ 6:01¿¡ "Jung Young Seok" <<a href="mailto:jung.youngseok@gmail.com" target="_blank">jung.youngseok@gmail.com</a>>´ÔÀÌ<br>
>ÀÛ¼º:<br>
><br>
><br>
>Dear GlusterFS Engineer,<br>
><br>
>I have questions that my glusterfs server and fuse client<br>
>perform properly on below specification.<br>
><br>
</div>>It can write only *65MB*/s through FUSE client to 1 glusterfs server (1<br>
<div>>brick and no replica for 1 volume )<br>
> - NW bandwidth are enough for now. I've check it with iftop<br>
</div>> - However it can write *120MB*/s when I mount nfs on the same volume.<br>
<div>><br>
>Could anyone check if the glusterfs and fuse client perform properly?<br>
><br>
><br>
>Detail explanations are below.<br>
>=======================================================================<br>
>I've set 4 glusterfs servers and 1 fuse client.<br>
>Each spec is as followings.<br>
><br>
</div>>*Server x 4*<br>
<div>> - CPU : Intel(R) Xeon(R) CPU E5-2609 0 @ 2.40GHz (2 cpu * 4 core)<br>
> - Memory : 32GB<br>
> - HDD (3TB 7.2K RPM SATA x 14 )<br>
> * RAID6(33T)<br>
> * XFS<br>
> - OS : RHS 2.1<br>
> - 4 Gluster Server will be used 2 replica x 2 distributed as 1 volume<br>
> - NW 1G for replica<br>
> - NW 1G for Storage and management<br>
<br>
</div>No need. The fuse client connects to all the servers. Replication happens from the client.<br>
<div><br>
> - Current active profile: rhs-high-throughput<br>
><br>
</div>>*FUSE Client (gluster 3.4)*<br>
<div>> - CPU : Intel(R) Xeon(R) CPU E5-2640 0 @ 2.50GHz<br>
> - Memory : 32GB<br>
> - OS : CentOS6.4<br>
> - NW 2G for Storage (NIC bonding)<br>
><br>
>All server will be in 10G network. (for now 1G network)<br>
><br>
><br>
>I've tested to check primitive disk performance.<br>
> - on first glusterfs server<br>
>* it can write 870MB/s (dd if=/dev/zero of=./dummy bs=4096 count=10000)<br>
> * it can read 1GB/s (cat test_file.23 > /dev/null )<br>
> - on fuse client (mount volume : 1 brick(1dist, no-replica)<br>
> * it can write 64.8MB/s<br>
> - on nfs client (mount volume : 1 brick(1dist, no-replica)<br>
> * it can write 120MB/s (it reached NW bandwith<br>
<br>
</div>My usual question here is how does dd represent your expected use case? Are you comparing apples to orchards?<br>
<div><br>
><br>
><br>
>I wonder why fuse client much slower than nfs client. (it's no-replica peer)<br>
>Is it normal performance?<br>
<br>
</div>I always max out my network connection with the fuse client, so no. It's not normal.<br>
<br>
<br></div></div>
_______________________________________________<br>
Gluster-users mailing list<br>
<a href="mailto:Gluster-users@gluster.org" target="_blank">Gluster-users@gluster.org</a><br>
<a href="http://supercolony.gluster.org/mailman/listinfo/gluster-users" target="_blank">http://supercolony.gluster.org/mailman/listinfo/gluster-users</a></blockquote></div><br></div>
</blockquote></div><br></div></div></div></div>