<div dir="ltr"><div><div>Hi, Pavel,<br></div>Thanks for your explanation.<br><br></div>- Hui<br><div><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Apr 28, 2015 at 11:45 AM, Pavel Emelyanov <span dir="ltr"><<a href="mailto:xemul@parallels.com" target="_blank">xemul@parallels.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On 04/28/2015 06:35 PM, Hui Kang wrote:<br>
<div><div class="h5">><br>
> On Tue, Apr 28, 2015 at 11:28 AM, Pavel Emelyanov <<a href="mailto:xemul@parallels.com">xemul@parallels.com</a> <mailto:<a href="mailto:xemul@parallels.com">xemul@parallels.com</a>>> wrote:<br>
><br>
> On 04/28/2015 06:23 PM, Hui Kang wrote:<br>
> > Hi,<br>
> > I have tested about the disk-less page migrations. As the instruction says, after dumping<br>
> > pages to the destination, there exists some small pages on the source node.<br>
> ><br>
> > I am just curious what is the rational of not transferring all the dumped pages. Thanks.<br>
><br>
> If you request for --page-server dump (or pre-dump) no pages are put in<br>
> the source images. Do you see non-empty pages*.img files?<br>
><br>
><br>
> I follow the instructions on this link (<a href="http://criu.org/Disk-less_migration" target="_blank">http://criu.org/Disk-less_migration</a>). Step 4 says "As was<br>
> said, CRIU images other than those with pages are very small and this won't take long. "<br>
> I am asking why those CRIU images are not sent to the page-server.<br>
<br>
</div></div>Ah, I see.<br>
<br>
The idea behind this is quite simple. In order to transfer images as they are generated requires<br>
a non-trivial protocol for it. We didn't want to implement it since there exists plenty of them.<br>
And sine images are typically several hundreds kilos in size, we can just dump them into tmpfs,<br>
then scp (or rsync or whatever) into the destination side. With pages this trick doesn't work for<br>
two reasons. First, pages images can occupy gigs, we can't put them on tmpfs. Second -- while<br>
doing iterations we need to drop the overwritten pages from images, thus the transferring protocol<br>
is not just "copy". So, as a balance, we implemented the protocol for pages only.<br>
<br>
If you have ideas how to improve that and transfer all images w/o intermediate storage (even tmpfs)<br>
you're welcome to share it :)<br>
<span class="HOEnZb"><font color="#888888"><br>
-- Pavel<br>
<br>
</font></span></blockquote></div><br></div></div></div></div></div>