[no subject]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Thanks
Bjoern
From: Peter Krempa <pkrempa@xxxxxxxxxx>
Date: Tuesday, April 19, 2022 at 6:20 AM
To: Bjoern Teipel <bjoern.teipel@xxxxxxxxxxxxx>
Cc: libvirt-users@xxxxxxxxxx <libvirt-users@xxxxxxxxxx>
Subject: Re: Virtio-scsi and block mirroring
CAUTION: This message originated externally, please use caution when clicki=
ng on links or opening attachments!


On Thu, Apr 14, 2022 at 16:36:38 +0000, Bjoern Teipel wrote:
> Hello everyone,

Hi,

>
> I=92m looking at an issue where I do see guests freezing (Dl) process sta=
te during a block disk mirror from one storage to another storage (NFS) whe=
re the network stack of the guest can freeze for up to 10 seconds.
> Looking at the storage and IO I noticed good throughput ad low latency <3=
ms and I am having trouble to track down the source for the issue, as neith=
er storage nor networking  show issues. Interestingly when I do the same te=
st with virtio-blk I do not really see the process freezes at the frequency=
 or duration compared to virtio-scsi which seem to indicate a client side r=
ather than storage side problem.

Hmm, this is really weird if the difference is in the guest-facing
device frontend.

Since libvirt is merely setting up the block job for the copy and the
copy itself is handled by qemu I suggest you contact the
qemu-block@xxxxxxxxxx mailing list.

Unfortunately you didn't provide any information on the disk
configuration (the VM XML) or how you start the blockjob, which I could
translate for you into qemu specifics. If you provide such information I
can do that to ensure that the qemu folks have all the relevant
information.

--_000_SA0PR20MB3502A1C05384B81CA52B0F01FBF29SA0PR20MB3502namp_
Content-Type: text/html; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable

<html xmlns:o=3D"urn:schemas-microsoft-com:office:office" xmlns:w=3D"urn:sc=
hemas-microsoft-com:office:word" xmlns:m=3D"http://schemas.microsoft.com/of=
fice/2004/12/omml" xmlns=3D"http://www.w3.org/TR/REC-html40";>
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3DWindows-1=
252">
<meta name=3D"Generator" content=3D"Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
=09{font-family:"Cambria Math";
=09panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
=09{font-family:Calibri;
=09panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
=09{margin:0in;
=09font-size:11.0pt;
=09font-family:"Calibri",sans-serif;}
.MsoChpDefault
=09{mso-style-type:export-only;
=09font-size:10.0pt;}
@page WordSection1
=09{size:8.5in 11.0in;
=09margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
=09{page:WordSection1;}
--></style>
</head>
<body lang=3D"EN-US" link=3D"#0563C1" vlink=3D"#954F72" style=3D"word-wrap:=
break-word">
<div class=3D"WordSection1">
<p class=3D"MsoNormal">Thanks Peter, that=92s what I figured that the actua=
l copy is done by the qemu process.<o:p></o:p></p>
<p class=3D"MsoNormal">The copy job is setup by openstack volume migration =
and translate into<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Courier New&quot;">=
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &lt;mirror type=3D'file' file=3D'/var/lib/no=
va/mnt/xxx' format=3D'raw' job=3D'copy'&gt;<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Courier New&quot;">=
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &lt;format type=3D'raw'/&gt;<o:p=
></o:p></span></p>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Courier New&quot;">=
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &lt;source file=3D'/var/lib/nova=
/mnt/yyy' index=3D'4'/&gt;<o:p></o:p></span></p>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Courier New&quot;">=
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &lt;backingStore/&gt;<o:p></o:p>=
</span></p>
<p class=3D"MsoNormal"><span style=3D"font-family:&quot;Courier New&quot;">=
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &lt;/mirror&gt;<o:p></o:p></span></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">From what I observed the issue is more noticeable wh=
en I see more fdatasync calls during the copy but I haven=92t been able to =
correlate that to the issue 100% yet<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Thanks<o:p></o:p></p>
<p class=3D"MsoNormal">Bjoern<o:p></o:p></p>
<div style=3D"border:none;border-top:solid #B5C4DF 1.0pt;padding:3.0pt 0in =
0in 0in">
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt"><b><span style=3D"fon=
t-size:12.0pt;color:black">From:
</span></b><span style=3D"font-size:12.0pt;color:black">Peter Krempa &lt;pk=
rempa@xxxxxxxxxx&gt;<br>
<b>Date: </b>Tuesday, April 19, 2022 at 6:20 AM<br>
<b>To: </b>Bjoern Teipel &lt;bjoern.teipel@xxxxxxxxxxxxx&gt;<br>
<b>Cc: </b>libvirt-users@xxxxxxxxxx &lt;libvirt-users@xxxxxxxxxx&gt;<br>
<b>Subject: </b>Re: Virtio-scsi and block mirroring<o:p></o:p></span></p>
</div>
<div>
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt">CAUTION: This message=
 originated externally, please use caution when clicking on links or openin=
g attachments!<br>
<br>
<br>
On Thu, Apr 14, 2022 at 16:36:38 +0000, Bjoern Teipel wrote:<br>
&gt; Hello everyone,<br>
<br>
Hi,<br>
<br>
&gt;<br>
&gt; I=92m looking at an issue where I do see guests freezing (Dl) process =
state during a block disk mirror from one storage to another storage (NFS) =
where the network stack of the guest can freeze for up to 10 seconds.<br>
&gt; Looking at the storage and IO I noticed good throughput ad low latency=
 &lt;3ms and I am having trouble to track down the source for the issue, as=
 neither storage nor networking&nbsp; show issues. Interestingly when I do =
the same test with virtio-blk I do not really
 see the process freezes at the frequency or duration compared to virtio-sc=
si which seem to indicate a client side rather than storage side problem.<b=
r>
<br>
Hmm, this is really weird if the difference is in the guest-facing<br>
device frontend.<br>
<br>
Since libvirt is merely setting up the block job for the copy and the<br>
copy itself is handled by qemu I suggest you contact the<br>
qemu-block@xxxxxxxxxx mailing list.<br>
<br>
Unfortunately you didn't provide any information on the disk<br>
configuration (the VM XML) or how you start the blockjob, which I could<br>
translate for you into qemu specifics. If you provide such information I<br=
>
can do that to ensure that the qemu folks have all the relevant<br>
information.<o:p></o:p></p>
</div>
</div>
</body>
</html>

--_000_SA0PR20MB3502A1C05384B81CA52B0F01FBF29SA0PR20MB3502namp_--




[Index of Archives]     [Virt Tools]     [Lib OS Info]     [Fedora Users]     [Fedora Desktop]     [Fedora SELinux]     [Yosemite News]     [KDE Users]

  Powered by Linux