Bug#864642: vmxnet3: Reports suspect GRO implementation on vSphere hosts / one VM crashes
forwarded #864642 https://bugzilla.kernel.org/show_bug.cgi?id=191201
thanks
Am 11.07.2017 um 10:24 schrieb Patrick Matthäi:
> found #864642 4.9.30-2+deb9u2
> thanks
>
> And it still crashes..
>
>
> Am 22.06.2017 um 14:58 schrieb Patrick Matthäi:
>> found #864642 4.9.30-2+deb9u1
>> thanks
>>
>>
>> Am 12.06.2017 um 10:02 schrieb Patrick Matthäi:
>>> Package: src:linux
>>> Version: 4.9.30-1
>>> Severity: important
>>> File: linux
>>>
>>> Dear Maintainer,
>>>
>>> *** Reporter, please consider answering these questions, where
>>> appropriate ***
>>>
>>> Since updating the kernel from linux-image-4.9.0-2-amd64 (4.9.18-1) to
>>> linux-image-4.9.0-3-amd64 (4.9.30-1) all VMs report - just for the
>>> "primary" interface this:
>>>
>>> TCP: ens192: Driver has suspect GRO implementation, TCP performance may
>>> be compromised.
>>>
>>> I can't see any performance impact. This happens on all our vSphere 6.0
>>> and 6.5 hosts (running on HPE ProLiant DL 360 G8 - G9 HW / ProLiant ML
>>> 350 G9 and so on).
>>>
>>> Why is this bug important? Because on one VM this also produces a kernel
>>> panic after some time (minutes or hours). I just could get the panic
>>> attached as screenshot. The only "big" difference between the crashing
>>> host and the others may be, that it is also running PM2, NPM, NodeJS and
>>> a NFS kernel server.
>>>
>>> If I boot the VM with 4.9.30-1 and deactivate gro and lro with:
>>> ethtool -K ens192 gro off
>>> ethtool -K ens192 lro off
>>> .. it does not crash.
>>>
>>> Booting 4.9.18-1 and everything is completly fine ;)
>>>
>> The VM keeps on crashing after a few hours
>>
--
/*
Mit freundlichem Gruß / With kind regards,
Patrick Matthäi
GNU/Linux Debian Developer
Blog: http://www.linux-dev.org/
E-Mail: pmatthaei@debian.org
patrick@linux-dev.org
*/
Reply to: