Bug 1138958

Summary: RFE Set MTU to 1400 in dnsmasq-neutron.conf if VXLAN or GRE are being installed
Product: Red Hat OpenStack Reporter: John Fulton <johfulto>
Component: openstack-packstackAssignee: Javier Peña <jpena>
Status: CLOSED ERRATA QA Contact: Itzik Brown <itbrown>
Severity: unspecified Docs Contact:
Priority: medium    
Version: 5.0 (RHEL 7)CC: afazekas, aortega, chale, derekh, dnavale, ichavero, johfulto, jpena, jschluet, jswensso, mmagr, myllynen, nyechiel, oblaut, tfreger, yeylon
Target Milestone: z5Keywords: FutureFeature, ZStream
Target Release: 5.0 (RHEL 7)   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: openstack-packstack-2014.1.1-0.47.dev1288.el7ost Doc Type: Enhancement
Doc Text:
With this update, a new feature in Packstack detects if VXLAN or GRE encapsulation is being used for OpenStack Networking tenant networks. If detected, it configures the DHCP agent to use a lower MTU for instances (1400 bytes). This is due to the fact that when using encapsulation, the VXLAN or GRE headers add some bytes to the Ethernet packet. If the instance MTU has the same size as the host MTU, some packets may be too large, resulting in slow network performance. As a result of this new feature, the lower MTU ensures the network traffic to instances performs adequately.
Story Points: ---
Clone Of:
: 1148044 (view as bug list) Environment:
Last Closed: 2015-09-10 11:36:27 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 1148044    

Description John Fulton 2014-09-06 18:39:27 UTC
Description of problem:

Customers who install Neutron with VXLAN or GRE get poor network performance until they set the MTU of their guests to 1400 as described in KCS 786743 (https://access.redhat.com/solutions/786743). Could the puppet modules check if VXLAN or GRE is going to be installed and then implement the resolution found in KCS 786743? 

Version-Release number of selected component (if applicable):

 openstack-puppet-modules-2014.1-19.3.el7ost.noarch

How reproducible:

Deterministic. 

Steps to Reproduce:

* Perform a packstack installation with a compute node separate from its network node which includes VXLAN or GRE. 
* Send network traffic from an instance on the compute node and observe log messages like "dropped over-mtu packet: 1550 > 1500"

Actual results:

* Guests on the compute node will have VXLAN or GRE overhead and some packets will be dropped because of the default MTU of 1500 on their NICs; resulting in poor network performance. 

Expected results:

* Guests on the compute node will have VXLAN or GRE overhead but packets will not be dropped for this reason because the extra byes will fit within default MTU of 1500 on the compute node's NICs because the guest MTU will be automatically set 1400; resulting in normal network performance. 

Additional info:

* This will save customers from having to deal with this problem after an installation and possibly reduce calls to GSS.

Comment 1 John Fulton 2014-09-06 18:39:50 UTC
*** Bug 1138954 has been marked as a duplicate of this bug. ***

Comment 3 Ivan Chavero 2014-09-26 03:43:16 UTC
The neutron puppet module has a network_device_mtu parameter i think this change should be done in packstack and the other installers that use the module to install and configure neutron.

Is this acceptable?

Comment 4 John Fulton 2014-09-30 15:00:04 UTC
Having packstack and other installers set the network_device_mtu parameter in puppet seems reasonable to me. I will update bugs.

Comment 8 Javier Peña 2014-11-04 14:13:41 UTC
(In reply to Ivan Chavero from comment #3)
> The neutron puppet module has a network_device_mtu parameter i think this
> change should be done in packstack and the other installers that use the
> module to install and configure neutron.
> 
> Is this acceptable?

I have done a test with network_device_mtu=1400. In this case, all ports created inside the network node netns use mtu=1400, however the guest mtu is still 1500. Would this be enough to fix the performance issue?

Comment 9 Javier Peña 2014-11-05 14:09:13 UTC
Just confirmed it is not enough, we need to set the guest MTU as specified in the KB article.

Comment 10 Javier Peña 2014-11-05 16:53:35 UTC
Proposed patch https://review.openstack.org/132888 for review.

Comment 12 Ivan Chavero 2015-08-13 01:17:41 UTC

-- 
Fedora Bugzappers volunteer triage team
https://fedoraproject.org/wiki/BugZappers

Comment 14 Itzik Brown 2015-08-31 09:19:35 UTC
Checked with version:
openstack-packstack-puppet-2014.1.1-0.47.dev1288.el7ost.noarch
openstack-packstack-2014.1.1-0.47.dev1288.el7ost.noarch

Comment 17 errata-xmlrpc 2015-09-10 11:36:27 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://rhn.redhat.com/errata/RHBA-2015-1746.html