Date: Wed, 28 Feb 2024 22:31:06 +0300 From: Vitaliy Gusev <gusev.vitaliy@gmail.com> To: Matthew Grooms <mgrooms@shrew.net> Cc: virtualization@freebsd.org Subject: Re: bhyve disk performance issue Message-ID: <1DAEB435-A613-4A04-B63F-D7AF7A0B7C0A@gmail.com> In-Reply-To: <25ddf43d-f700-4cb5-af2a-1fe669d1e24b@shrew.net> References: <6a128904-a4c1-41ec-a83d-56da56871ceb@shrew.net> <28ea168c-1211-4104-b8b4-daed0e60950d@app.fastmail.com> <0ff6f30a-b53a-4d0f-ac21-eaf701d35d00@shrew.net> <6f6b71ac-2349-4045-9eaf-5c50d42b89be@shrew.net> <50614ea4-f0f9-44a2-b5e6-ebb33cfffbc4@shrew.net> <6a4e7e1d-cca5-45d4-a268-1805a15d9819@shrew.net> <f01a9bca-7023-40c0-93f2-8cdbe4cd8078@tubnor.net> <edb80fff-561b-4dc5-95ee-204e0c6d95df@shrew.net> <a07d070b-4dc1-40c9-bc80-163cd59a5bfc@Duedinghausen.eu> <e45c95df-4858-48aa-a274-ba1bf8e599d5@shrew.net> <BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com> <25ddf43d-f700-4cb5-af2a-1fe669d1e24b@shrew.net>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --]
Hi, Matthew.
I still do not know what command line was used for bhyve. I couldn't find it through the thread, sorry. And I couldn't find virtual disk size that you used.
Could you, please, simplify bonnie++ output, it is hard to decode due to alignment and use exact numbers for:
READ seq - I see you had 1.6GB/s for the good time and ~500MB/s for the worst.
WRITE seq - ...
If you have slow results both for the read and write operations, you probably should perform testing only for READs and do not do anything until READs are fine.
Again, if you have slow performance for Ext4 Filesystem in guest VM placed on the passed disk image, you should try to test on the raw disk image, i.e. without Ext4, because it could be related.
If you run test inside VM on a filesystem, you can have deal with filesystem bottlenecks, bugs, fragmentation etc. Do you want to fix them all? I don’t think so.
For example, if you pass disk image 40G and create Ext4 filesystem, and during testing the filesystem becomes full over 80%, I/O could be performed not so fine.
You probably should eliminate that guest filesystem behaviour when you meet IO performance slowdown.
Also, please look at the TRIM operations when you perform WRITE testing. It could be also related to the slow write I/O.
——
Vitaliy
> On 28 Feb 2024, at 21:29, Matthew Grooms <mgrooms@shrew.net> wrote:
>
> On 2/27/24 04:21, Vitaliy Gusev wrote:
>> Hi,
>>
>>
>>> On 23 Feb 2024, at 18:37, Matthew Grooms <mgrooms@shrew.net> <mailto:mgrooms@shrew.net> wrote:
>>>
>>>> ...
>>> The problem occurs when an image file is used on either ZFS or UFS. The problem also occurs when the virtual disk is backed by a raw disk partition or a ZVOL. This issue isn't related to a specific underlying filesystem.
>>>
>>
>> Do I understand right, you ran testing inside VM inside guest VM on ext4 filesystem? If so you should be aware about additional overhead in comparison when you were running tests on the hosts.
>>
> Hi Vitaliy,
>
> I appreciate you providing the feedback and suggestions. I spent over a week trying as many combinations of host and guest options as possible to narrow this issue down to a specific host storage or a guest device model option. Unfortunately the problem occurred with every combination I tested while running Linux as the guest. Note, I only tested RHEL8 & RHEL9 compatible distributions ( Alma & Rocky ). The problem did not occur when I ran FreeBSD as the guest. The problem did not occur when I ran KVM in the host and Linux as the guest.
>
>> I would suggest to run fio (or even dd) on raw disk device inside VM, i.e. without filesystem at all. Just do not forget do “echo 3 > /proc/sys/vm/drop_caches” in Linux Guest VM before you run tests.
> The two servers I was using to test with are are no longer available. However, I'll have two more identical servers arriving in the next week or so. I'll try to run additional tests and report back here. I used bonnie++ as that was easily installed from the package repos on all the systems I tested.
>
>>
>> Could you also give more information about:
>>
>> 1. What results did you get (decode bonnie++ output)?
> If you look back at this email thread, there are many examples of running bonnie++ on the guest. I first ran the tests on the host system using Linux + ext4 and FreeBSD 14 + UFS & ZFS to get a baseline of performance. Then I ran bonnie++ tests using bhyve as the hypervisor and Linux & FreeBSD as the guest. The combination of host and guest storage options included ...
>
> 1) block device + virtio blk
> 2) block device + nvme
> 3) UFS disk image + virtio blk
> 4) UFS disk image + nvme
> 5) ZFS disk image + virtio blk
> 6) ZFS disk image + nvme
> 7) ZVOL + virtio blk
> 8) ZVOL + nvme
>
> In every instance, I observed the Linux guest disk IO often perform very well for some time after the guest was first booted. Then the performance of the guest would drop to a fraction of the original performance. The benchmark test was run every 5 or 10 minutes in a cron job. Sometimes the guest would perform well for up to an hour before performance would drop off. Most of the time it would only perform well for a few cycles ( 10 - 30 mins ) before performance would drop off. The only way to restore the performance was to reboot the guest. Once I determined that the problem was not specific to a particular host or guest storage option, I switched my testing to only use a block device as backing storage on the host to avoid hitting any system disk caches.
>
> Here is the test script I used in the cron job ...
>
> #!/bin/sh
> FNAME='output.txt'
>
> echo ================================================================================ >> $FNAME
> echo Begin @ `/usr/bin/date` >> $FNAME
> echo >> $FNAME
> /usr/sbin/bonnie++ 2>&1 | /usr/bin/grep -v 'done\|,' >> $FNAME
> echo >> $FNAME
> echo End @ `/usr/bin/date` >> $FNAME
>
> As you can see, I'm calling bonnie++ with the system defaults. That uses a data set size that's 2x the guest RAM in an attempt to minimize the effect of filesystem cache on results. Here is an example of the output that bonnie++ produces ...
>
> Version 2.00 ------Sequential Output------ --Sequential Input- --Random-
> -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
> Name:Size etc /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
> linux-blk 63640M 694k 99 1.6g 99 737m 76 985k 99 1.3g 69 +++++ +++
> Latency 11579us 535us 11889us 8597us 21819us 8238us
> Version 2.00 ------Sequential Create------ --------Random Create--------
> linux-blk -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
> files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
> 16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++
> Latency 7620us 126us 1648us 151us 15us 633us
>
> --------------------------------- speed drop ---------------------------------
>
> Version 2.00 ------Sequential Output------ --Sequential Input- --Random-
> -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
> Name:Size etc /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
> linux-blk 63640M 676k 99 451m 99 314m 93 951k 99 402m 99 15167 530
> Latency 11902us 8959us 24711us 10185us 20884us 5831us
> Version 2.00 ------Sequential Create------ --------Random Create--------
> linux-blk -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
> files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP
> 16 0 96 +++++ +++ +++++ +++ 0 96 +++++ +++ 0 75
> Latency 343us 165us 1636us 113us 55us 1836us
>
> In the example above, the benchmark test repeated about 20 times with results that were similar to the performance shown above the dotted line ( ~ 1.6g/s seq write and 1.3g/s seq read ). After that, the performance dropped to what's shown below the dotted line which is less than 1/4 the original speed ( ~ 451m/s seq write and 402m/s seq read ).
>
>> 2. What results expecting?
>>
> What I expect is that, when I perform the same test with the same parameters, the results would stay more or less consistent over time. This is true when KVM is used as the hypervisor on the same hardware and guest options. That said, I'm not worried about bhyve being consistently slower than kvm or a FreeBSD guest being consistently slower than a Linux guest. I'm concerned that the performance drop over time is indicative of an issue with how bhyve interacts with non-freebsd guests.
>
>> 3. VM configuration, virtio-blk disk size, etc.
>> 4. Full command for tests (including size of test-set), bhyve, etc.
> I believe this was answered above. Please let me know if you have additional questions.
>
>>
>> 5. Did you pass virtio-blk as 512 or 4K ? If 512, probably you should try 4K.
>>
> The testing performed was not exclusively with virtio-blk.
>
>
>> 6. Linux has several read-ahead options for IO schedule, and it could be related too.
>>
> I suppose it's possible that bhyve could be somehow causing the disk scheduler in the Linux guest to act differently. I'll see if I can figure out how to disable that in future tests.
>
>
>> Additionally could also you play with “sync=disabled” volume/zvol option? Of course it is only for write testing.
> The testing performed was not exclusively with zvols.
>
>
> Once I have more hardware available, I'll try to report back with more testing. It may be interesting to also see how a Windows guest performs compared to Linux & FreeBSD. I suspect that this issue may only be triggered when a fast disk array is in use on the host. My tests use a 16x SSD RAID 10 array. It's also quite possible that the disk IO slowdown is only a symptom of another issue that's triggered by the disk IO test ( please see end of my last post related to scheduler priority observations ). All I can say for sure is that ...
>
> 1) There is a problem and it's reproducible across multiple hosts
> 2) It affects RHEL8 & RHEL9 guests but not FreeBSD guests
> 3) It is not specific to any host or guest storage option
>
> Thanks,
>
> -Matthew
>
[-- Attachment #2 --]
<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body style="overflow-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;"><span style="font-size: 15px;">Hi, Matthew.</span><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">I still do not know what command line was used for bhyve. I couldn't find it through the thread, sorry. And I </span><span style="font-size: 15px;">couldn't find virtual disk size that you used.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">Could you, please, simplify bonnie++ output, it is hard to decode due to alignment and use exact numbers for:</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">READ seq - I see you had 1.6GB/s for the good time and ~500MB/s for the worst.</span></div><div><span style="font-size: 15px;">WRITE seq - ...</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">If you have slow results both for the read and write operations, you probably should perform testing <u>only</u> for READs and do not do anything until READs are fine.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">Again, if you have slow performance for Ext4 Filesystem in guest VM placed on the passed disk image, you should </span><span style="font-size: 15px;">try to test on the raw disk image, i.e. without Ext4, because it could be related.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">If you run test inside VM on a filesystem, you can have deal with filesystem bottlenecks, bugs, fragmentation etc. Do you want to fix them all? I don’t think so.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">For example, if you pass disk image 40G and create Ext4 filesystem, and during testing the filesystem becomes full over 80%, I/O could be performed not so fine.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">You probably should eliminate that guest filesystem behaviour when you meet IO performance slowdown.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">Also, please look at the TRIM operations when you perform WRITE testing. It could be also related to the slow write I/O.</span></div><div><span style="font-size: 15px;"><br></span></div><div><span style="font-size: 15px;">——</span></div><div><span style="font-size: 15px;">Vitaliy</span></div><div><div><br><blockquote type="cite"><div>On 28 Feb 2024, at 21:29, Matthew Grooms <mgrooms@shrew.net> wrote:</div><br class="Apple-interchange-newline"><div>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<div>
<div class="moz-cite-prefix">On 2/27/24 04:21, Vitaliy Gusev wrote:<br>
</div>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
Hi,
<div><br>
</div>
<div>
<div><br>
<blockquote type="cite">
<div>On 23 Feb 2024, at 18:37, Matthew Grooms
<a class="moz-txt-link-rfc2396E" href="mailto:mgrooms@shrew.net"><mgrooms@shrew.net></a> wrote:</div>
<br class="Apple-interchange-newline">
<div>
<div>
<blockquote type="cite">...</blockquote>
The problem occurs when an image file is used on either
ZFS or UFS. The problem also occurs when the virtual
disk is backed by a raw disk partition or a ZVOL. This
issue isn't related to a specific underlying filesystem.<br>
<br>
</div>
</div>
</blockquote>
<div><br>
</div>
Do I understand right, you ran testing inside VM inside guest
VM on ext4 filesystem? If so you should be aware about
additional overhead in comparison when you were running tests
on the hosts.</div>
<div><br>
</div>
</div>
</blockquote><p>Hi Vitaliy,<br>
<br>
I appreciate you providing the feedback and suggestions. I spent
over a week trying as many combinations of host and guest options
as possible to narrow this issue down to a specific host storage
or a guest device model option. Unfortunately the problem occurred
with every combination I tested while running Linux as the guest.
Note, I only tested RHEL8 & RHEL9 compatible distributions (
Alma & Rocky ). The problem did not occur when I ran FreeBSD
as the guest. The problem did not occur when I ran KVM in the host
and Linux as the guest.<br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div>I would suggest to run fio (or even dd) on raw disk device
inside VM, i.e. without filesystem at all. Just do not forget
do “<span style="font-family: Menlo; font-size: 12px; background-color: rgb(231, 238, 238);">echo
3 > /proc/sys/vm/drop_caches</span>” in Linux Guest VM
before you run tests. <br>
</div>
</div>
</blockquote><p>The two servers I was using to test with are are no longer
available. However, I'll have two more identical servers arriving
in the next week or so. I'll try to run additional tests and
report back here. I used bonnie++ as that was easily installed
from the package repos on all the systems I tested.<br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div><br>
</div>
<div>Could you also give more information about:</div>
<div><br>
</div>
<div> 1. What results did you get (decode bonnie++ output)?</div>
</div>
</blockquote><p>If you look back at this email thread, there are many examples of
running bonnie++ on the guest. I first ran the tests on the host
system using Linux + ext4 and FreeBSD 14 + UFS & ZFS to get a
baseline of performance. Then I ran bonnie++ tests using bhyve as
the hypervisor and Linux & FreeBSD as the guest. The
combination of host and guest storage options included ...<br>
<br>
1) block device + virtio blk<br>
2) block device + nvme<br>
3) UFS disk image + virtio blk<br>
4) UFS disk image + nvme<br>
5) ZFS disk image + virtio blk<br>
6) ZFS disk image + nvme<br>
7) ZVOL + virtio blk<br>
8) ZVOL + nvme<br>
<br>
In every instance, I observed the Linux guest disk IO often
perform very well for some time after the guest was first booted.
Then the performance of the guest would drop to a fraction of the
original performance. The benchmark test was run every 5 or 10
minutes in a cron job. Sometimes the guest would perform well for
up to an hour before performance would drop off. Most of the time
it would only perform well for a few cycles ( 10 - 30 mins )
before performance would drop off. The only way to restore the
performance was to reboot the guest. Once I determined that the
problem was not specific to a particular host or guest storage
option, I switched my testing to only use a block device as
backing storage on the host to avoid hitting any system disk
caches.<br>
<br>
Here is the test script I used in the cron job ...<br>
<br>
<font size="2" face="monospace">#!/bin/sh<br>
FNAME='output.txt'<br>
</font>
<font face="monospace"><font size="2"><br>
echo
================================================================================
>> $FNAME<br>
echo Begin @ `/usr/bin/date` >> $FNAME<br>
echo >> $FNAME<br>
/usr/sbin/bonnie++ 2>&1 | /usr/bin/grep -v 'done\|,'
>> $FNAME<br>
echo >> $FNAME<br>
echo End @ `/usr/bin/date` >> $FNAME</font><br>
</font>
<br>
As you can see, I'm calling bonnie++ with the system defaults.
That uses a data set size that's 2x the guest RAM in an attempt to
minimize the effect of filesystem cache on results. Here is an
example of the output that bonnie++ produces ...<br>
<br>
<font face="monospace"><font size="2">Version 2.00
------Sequential Output------ --Sequential Input- --Random-<br>
-Per Chr- --Block-- -Rewrite- -Per Chr-
--Block-- --Seeks--<br>
Name:Size etc /sec %CP /sec %CP /sec %CP /sec %CP
/sec %CP /sec %CP<br>
linux-blk 63640M 694k 99 1.6g 99 737m 76 985k 99
1.3g 69 +++++ +++<br>
Latency 11579us 535us 11889us 8597us
21819us 8238us<br>
Version 2.00 ------Sequential Create------
--------Random Create--------<br>
linux-blk -Create-- --Read--- -Delete-- -Create--
--Read--- -Delete--<br>
files /sec %CP /sec %CP /sec %CP /sec %CP
/sec %CP /sec %CP<br>
16 +++++ +++ +++++ +++ +++++ +++ +++++ +++
+++++ +++ +++++ +++<br>
Latency 7620us 126us 1648us 151us
15us 633us<br>
<br>
--------------------------------- speed drop
---------------------------------<br>
<br>
Version 2.00 ------Sequential Output------ --Sequential
Input- --Random-<br>
-Per Chr- --Block-- -Rewrite- -Per Chr-
--Block-- --Seeks--<br>
Name:Size etc /sec %CP /sec %CP /sec %CP /sec %CP
/sec %CP /sec %CP<br>
linux-blk 63640M 676k 99 451m 99 314m 93 951k 99
402m 99 15167 530<br>
Latency 11902us 8959us 24711us 10185us
20884us 5831us<br>
Version 2.00 ------Sequential Create------
--------Random Create--------<br>
linux-blk -Create-- --Read--- -Delete-- -Create--
--Read--- -Delete--<br>
files /sec %CP /sec %CP /sec %CP /sec %CP
/sec %CP /sec %CP<br>
16 0 96 +++++ +++ +++++ +++ 0 96
+++++ +++ 0 75<br>
Latency 343us 165us 1636us 113us
55us 1836us<br>
</font></font><br>
In the example above, the benchmark test repeated about 20 times
with results that were similar to the performance shown above the
dotted line ( ~ 1.6g/s seq write and 1.3g/s seq read ). After
that, the performance dropped to what's shown below the dotted
line which is less than 1/4 the original speed ( ~ 451m/s seq
write and 402m/s seq read ). <br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div> 2. What results expecting?<br>
<br>
</div>
</div>
</blockquote><p>What I expect is that, when I perform the same test with the same
parameters, the results would stay more or less consistent over
time. This is true when KVM is used as the hypervisor on the same
hardware and guest options. That said, I'm not worried about bhyve
being consistently slower than kvm or a FreeBSD guest being
consistently slower than a Linux guest. I'm concerned that the
performance drop over time is indicative of an issue with how
bhyve interacts with non-freebsd guests.<br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div> 3. VM configuration, virtio-blk disk size, etc.</div>
<div> 4. Full command for tests (including size of test-set),
bhyve, etc.<br>
</div>
</div>
</blockquote><p>I believe this was answered above. Please let me know if you have
additional questions.<br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div><br>
</div>
<div> 5. Did you pass virtio-blk as 512 or 4K ? If 512, probably
you should try 4K.<br>
<br>
</div>
</div>
</blockquote><p>The testing performed was not exclusively with virtio-blk.<br>
<br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div> 6. Linux has several read-ahead options for IO schedule,
and it could be related too.</div>
<div><br>
</div>
</div>
</blockquote><p>I suppose it's possible that bhyve could be somehow causing the
disk scheduler in the Linux guest to act differently. I'll see if
I can figure out how to disable that in future tests.<br>
<br>
</p>
<blockquote type="cite" cite="mid:BE794E98-7B69-4626-BB66-B56F23D6A67E@gmail.com">
<div>
<div>Additionally could also you play with “sync=disabled”
volume/zvol option? Of course it is only for write testing.<br>
</div>
</div>
</blockquote><p>The testing performed was not exclusively with zvols.<br>
<br>
</p><p>Once I have more hardware available, I'll try to report back with
more testing. It may be interesting to also see how a Windows
guest performs compared to Linux & FreeBSD. I suspect that
this issue may only be triggered when a fast disk array is in use
on the host. My tests use a 16x SSD RAID 10 array. It's also quite
possible that the disk IO slowdown is only a symptom of another
issue that's triggered by the disk IO test ( please see end of my
last post related to scheduler priority observations ). All I can
say for sure is that ...<br>
<br>
1) There is a problem and it's reproducible across multiple hosts<br>
2) It affects RHEL8 & RHEL9 guests but not FreeBSD guests<br>
3) It is not specific to any host or guest storage option<br>
<br>
Thanks,</p><p>-Matthew<br>
</p>
</div>
</div></blockquote></div><br></div></body></html>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1DAEB435-A613-4A04-B63F-D7AF7A0B7C0A>