RHEL Engineering is moving the tracking of its product development work on RHEL 6 through RHEL 9 to Red Hat Jira (issues.redhat.com). If you're a Red Hat customer, please continue to file support cases via the Red Hat customer portal. If you're not, please head to the "RHEL project" in Red Hat Jira and file new tickets here. Individual Bugzilla bugs in the statuses "NEW", "ASSIGNED", and "POST" are being migrated throughout September 2023. Bugs of Red Hat partners with an assigned Engineering Partner Manager (EPM) are migrated in late September as per pre-agreed dates. Bugs against components "kernel", "kernel-rt", and "kpatch" are only migrated if still in "NEW" or "ASSIGNED". If you cannot log in to RH Jira, please consult article #7032570. That failing, please send an e-mail to the RH Jira admins at rh-issues@redhat.com to troubleshoot your issue as a user management inquiry. The email creates a ServiceNow ticket with Red Hat. Individual Bugzilla bugs that are migrated will be moved to status "CLOSED", resolution "MIGRATED", and set with "MigratedToJIRA" in "Keywords". The link to the successor Jira issue will be found under "Links", have a little "two-footprint" icon next to it, and direct you to the "RHEL project" in Red Hat Jira (issue links are of type "https://issues.redhat.com/browse/RHEL-XXXX", where "X" is a digit). This same link will be available in a blue banner at the top of the page informing you that that bug has been migrated.
Bug 1465803 - cannot set MTU to VF
Summary: cannot set MTU to VF
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Red Hat Enterprise Linux 7
Classification: Red Hat
Component: NetworkManager
Version: 7.4
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: rc
: ---
Assignee: sushil kulkarni
QA Contact: Desktop QE
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2017-06-28 09:03 UTC by Vladimir Benes
Modified: 2017-07-13 16:15 UTC (History)
9 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2017-07-13 16:15:26 UTC
Target Upstream Version:
Embargoed:

Attachments (Terms of Use)

Description Vladimir Benes 2017-06-28 09:03:01 UTC 
Description of problem:
I have a bunch of VF devices and trying to set mtu 9000 on them like this:

nmcli connection add type ethernet ifname enp5s16 con-name ethie ipv4.method manual ipv4.address 1.2.3.4/24 802-3-ethernet.mtu 9000 

but after upping a connection I see:

enp5s16f2: connected to ethie
	"Intel 82599 Ethernet Controller Virtual Function"
	ethernet (ixgbevf), F6:6B:15:A3:20:F8, hw, mtu 1500
	inet4 1.2.3.4/24
	inet6 fe80::ff5a:9864:5170:786c/64

This seems to work incorrectly

Version-Release number of selected component (if applicable):
1.8.0-9
Comment 1 Thomas Haller 2017-06-28 09:47:57 UTC 
any log files available, or a beaker machine?
Comment 2 Vladimir Benes 2017-06-28 10:05:52 UTC 
Machine provided.
Comment 4 Thomas Haller 2017-06-28 14:22:13 UTC 
on the machine, I see:

Jun 28 10:11:20 wsfd-netdev5.lab.bos.redhat.com NetworkManager[2316]: <debug> [1498659080.6933] platform-linux: link: change 12: mtu: 9000
Jun 28 10:11:20 wsfd-netdev5.lab.bos.redhat.com NetworkManager[2316]: <debug> [1498659080.7006] platform-linux: do-request-link: 12
Jun 28 10:11:20 wsfd-netdev5.lab.bos.redhat.com NetworkManager[2316]: <trace> [1498659080.7006] platform-linux: event-notification: NEWLINK, seq 0: 12: enp5s16f1 <UP;broadcast,multicast,up,running> mtu 1500 arp 1 ethernet? not-init addrgenmode none addr 6E:DC:AF:20:A4:3D rx:0,0 tx:0,0
Jun 28 10:11:20 wsfd-netdev5.lab.bos.redhat.com NetworkManager[2316]: <debug> [1498659080.7006] platform-linux: netlink: recvmsg: error message from kernel: Invalid argument (22) for request 83
Jun 28 10:11:20 wsfd-netdev5.lab.bos.redhat.com NetworkManager[2316]: <trace> [1498659080.7007] platform-linux: event-notification: NEWLINK, seq 84: 12: enp5s16f1 <UP;broadcast,multicast,up,running> mtu 1500 arp 1 ethernet? not-init addrgenmode none addr 6E:DC:AF:20:A4:3D rx:0,0 tx:0,0
Jun 28 10:11:20 wsfd-netdev5.lab.bos.redhat.com NetworkManager[2316]: <error> [1498659080.7007] platform-linux: do-change-link[12]: failure changing link: failure 22 (Invalid argument)



likewise:

 # echo 2 > /sys/class/net/enp5s0f1/device/sriov_numvfs
 # ip link set enp5s16f1 mtu 9000
 RTNETLINK answers: Invalid argument


it seems like a driver limitation...

(needs more investigation).
Comment 5 Thomas Haller 2017-06-28 14:40:17 UTC 
ok, seems that the MTU of the PF must also be high enough.

iproute2 allows setting the MTU of the VF, if the underlying device has it set to 9000.
Interestingly, I can set the MTU of the VF to up to 9710 (if the PF has it at 9000).


The workaround seems to be to also activate a connection on the other interface with a sufficiently high MTU.

NM should somehow mediate the MTU of the VFs and the PF. Which, is however complicated to do.
Comment 6 Thomas Haller 2017-06-28 16:01:39 UTC 
I think it's too complicated for NM to figure out the correct MTU of the interface. It's up to the user to configure the maximum allowed MTU on the PF.

Also, because large MTU possibly needs configuration on the switch side. So, the MTU is not something chosen lightly, the user must ensure that the MTU fits into the larger picture.

I tend to close this as WONTFIX.
Comment 7 Vladimir Benes 2017-06-28 19:10:58 UTC 
Can we at least document this somehow?
Comment 8 Mirek Jahoda 2017-06-28 19:41:46 UTC 
Are you thinking about adding this BZ as Known Issue in the RHEL 7.4 Release Notes or about something else?
Comment 9 Vladimir Benes 2017-06-28 19:55:18 UTC 
I would more likely see some note about setting MTU in documentation for VFs and relation to PF. Should we add it under bug 1398934 ?
Comment 10 Mirek Jahoda 2017-06-29 13:48:01 UTC 
BZ#1398934 is also not included in the RHEL 7.4 Release Notes. Sushil, should I add this one? If yes, provide me please some Doc Text draft as a base for description (including Vladimir's suggestion for an additional note). Thank you.

(In reply to Vladimir Benes from comment #9)
> I would more likely see some note about setting MTU in documentation for VFs
> and relation to PF. Should we add it under bug 1398934 ?
Comment 11 sushil kulkarni 2017-06-29 15:39:32 UTC 
Sure.. I think 1398934 should be documented.. and we can add the additional note about the underlying PF configuration dependency.

Ben would you agree?
Comment 12 Beniamino Galvani 2017-06-29 16:38:15 UTC 
(In reply to sushil kulkarni from comment #11)
> Sure.. I think 1398934 should be documented.. and we can add the additional
> note about the underlying PF configuration dependency.
> 
> Ben would you agree?

Yes, I've added the documentation with a note about the MTU.
Note You need to log in before you can comment on or make changes to this bug.