<div>Hi Ryo<br></div>
<div>I attached new file that includes I/O total bandwidth of evaluation system.</div>
<div>We tested total bandwidth of weight policy by I/O in Dom0 and DomU system</div>
<div>and it is measured through iostat tool and dm-ioband debug patch which I gave you several months ago.</div>
<div>Of course, the result in prior report was measured by dm-ioband debug patch.</div>
<div> </div>
<div>As a result, the big difference in prior report derives from the location where we measured I/O bandwidth</div>
<div>iostat counts it in application level and dm-ioband debug patch does it in dm-ioband controller.</div>
<div>I think the difference is related with buffer cache.</div>
<div> </div>
<div>Thank you.</div>
<div>Have a nice weekend</div>
<div> </div>
<div class="gmail_quote">2009/8/27 Dong-Jae Kang <span dir="ltr"><<a href="mailto:baramsori72@gmail.com" target="_blank">baramsori72@gmail.com</a>></span><br>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Hi Ryo<br><br>
<div class="gmail_quote">2009/8/27 Ryo Tsuruta <span dir="ltr"><<a href="mailto:ryov@valinux.co.jp" target="_blank">ryov@valinux.co.jp</a>></span>
<div>
<div></div>
<div><br>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Hi Dong-Jae,<br><br># I've added dm-devel to Cc:.<br>
<div><br>Dong-Jae Kang <<a href="mailto:baramsori72@gmail.com" target="_blank">baramsori72@gmail.com</a>> wrote:<br>> Hi Ryo<br>><br>> I attached new test result file(ioband-partition-based-evaluation.xls)in<br>
> this mail.<br><br></div>Thanks for your great job.<br>
<div><br>> In this time, it is not virtualization environment.<br>> I evaluated partition-based use cases before I do it in vitualization<br>> environment.<br>> because I think the two cases are smilar each other.<br>
><br>> The detailed information about the evaluation can be referred in attached<br>> file.<br>><br>> If you have any questions or comments after examine it,<br>> please give me your opinion.<br><br></div>
I would like to know the throughput without dm-ioband in your<br>environment. Because the total throughput of range-bw policy is<br>8000KB/s, which means the device has a capability to perform over<br>8000KB/s, but the total throughput of weight policy is lower than<br>
the range-bw policy. In my environment, there is no significant<br>difference in average throughput between with and without dm-ioband.<br>I ran fio in the way described in your result file. Here are the<br>results of my environment. The throughputs were calculated from<br>
"iostat -k 1" outputs.<br><br> buffered write test<br> Avg. throughput [KB/s]<br> w/o ioband w/ioband<br>sdb2 14485 5788<br>sdb3 12494 22295<br>total 26979 28030<br>
</blockquote>
<div> </div></div></div>
<div>OK, good comments.</div>
<div>I omitted the total bandwidth of the evaluation system.</div>
<div> </div>
<div>I will reply to you about it tomorrow after I check and re-test it again.</div>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid"><span></span><br>Thanks,<br><font color="#888888">Ryo Tsuruta<br></font></blockquote></div><br>Thank you for comments.
<div>
<div></div>
<div><br clear="all">
<div></div><br>-- <br>Best Regards,<br>Dong-Jae Kang<br></div></div></blockquote></div><br><br clear="all">
<div></div><br>-- <br>Best Regards,<br>Dong-Jae Kang<br>