Bug 47112 – By Using Virt-IO Driver in Windows Packet Loss

Bug 47112 - By Using Virt-IO Driver in Windows Packet Loss


Summary:	By Using Virt-IO Driver in Windows Packet Loss

Status:	CLOSED FIXED

Product:	UCS
Classification:	Unclassified
Component:	Virtualization - KVM
Version:	UCS 4.2
Hardware:	Other other

Importance:	P5 major (vote)
Target Milestone:	UCS 4.2-4-errata
Assigned To:	Erik Damrose
QA Contact:	Philipp Hahn

URL:
Keywords:

Depends on:
Blocks:	47321 48360
	Show dependency tree / graph

Reported:	2018-06-01 10:42 CEST by Christian Völker
Modified:	2018-12-17 10:09 CET (History)
CC List:	2 users (show)

See Also:
What kind of report is it?:	Bug Report
What type of bug is this?:	5: Major Usability: Impairs usability in key scenarios
Who will be affected by this bug?:	3: Will affect average number of installed domains
How will those affected feel about the bug?:	3: A User would likely not purchase the product
User Pain:	0.257
Enterprise Customer affected?:	Yes
School Customer affected?:
ISV affected?:
Waiting Support:	Yes
Flags outvoted (downgraded) after PO Review:
Ticket number:	2018053021000228
Bug group (optional):
Max CVSS v3 score:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.

Description Christian Völker

2018-06-01 10:42:45 CEST

Customer is using VirtIO driver inside virtual Windows 2012R2 Machine.

After a while he recons massive packet losses on connections to this server.

An upgrade to the latest stable version 141 (from https://fedorapeople.org/groups/virt/virtio-win/direct-downloads/archive-virtio/virtio-win-0.1.141-1/virtio-win-0.1.141.iso ) did not change the behavior for long.

As soon as using the e1000 driver the issue goes away.

There exists an possibly related bug with the same symptoms:
https://bugzilla.redhat.com/show_bug.cgi?id=1451978


Customer used v105 and tested v141 with the same results.

Comment 2 Erik Damrose

2018-07-09 14:13:50 CEST

Commits for UCS 4.3 were wrongly tagged with this bug number, the UCS 4.3 bug is bug 47321

This will update the virtio driver to the current latest stable version. I could not reproduce the linked redhat issue with this driver version.

Driver date: Fri Aug 04 2017
Version shown for netkvm.sys: 62.74.104.14100

bd34dec5 Update to virtio driver 0.1.141-1
b985a195 yaml

Comment 3 Christian Völker

2018-07-25 07:50:53 CEST

In the initial description customer had tried 141-1.
Which did not work.

Fix is now to update to 141-1 (comment #2)?

Comment 4 Erik Damrose

2018-07-31 16:36:57 CEST

We discussed this internally. We can not currently reproduce the customer issue that was mentioned at the redhat bug. As at least the last stable driver version fixes that issue, we will ship the newest driver version with this bugfix.

If the customer issue persists, it has to be debugged more in depth directly in their environment.

Comment 5 Philipp Hahn

2018-08-07 13:35:13 CEST

OK: https://fedorapeople.org/groups/virt/virtio-win/CHANGELOG

OK: Win7
OK: Win8
OK: Win10
OK: Win2k12R2

OK: NetKVM
OK: vIOstor

OK: Live migration
OK: ping -I0

OK: errata-announce -V --only univention-kvm-virtio.yaml
OK: univention-kvm-virtio.yaml
OK: dpkg -c univention-kvm-virtio_6.0.0-3A~4.2.0.201807091407_all.deb

If the problem persists ask the customer to give <https://fedorapeople.org/groups/virt/virtio-win/direct-downloads/archive-virtio/virtio-win-0.1.149-2/virtio-win_amd64.vfd> a try.

Comment 6 Arvid Requate

2018-08-08 14:53:08 CEST

<http://errata.software-univention.de/ucs/4.2/428.html>