[openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

A, Keshava keshava.a at hp.com
Tue Oct 28 17:27:18 UTC 2014


Hi Alan and  Salvatore,

Thanks for response and I also agree we need to take small steps.
However I have below points to make.

It is very important how the Service VM needs will be deployed w.r.t HA.
As per current discussion, you are proposing something like below kind of deployment for Carrier Grade HA.
Since there is a separate port for Standby-VM also, then the corresponding standby-VM interface address should be globally routable also.
Means it may require the Standby Routing protocols to advertise its interface as Next-HOP for prefix it routes.
However external world should not be aware of the standby-routing running in the network.

[cid:image004.png at 01CFF302.7E2F7FC0]

[cid:image007.png at 01CFF301.6EB2C440]

Instead if we can think of running Standby on same stack with Passive port, ( as shown below)  then external world will be unaware of the standing Service Routing running.
This may be  something very basic requirement from Service-VM (NFV HA perspective) for Routing/MPLS/Packet processing domain.
I am brining this issue now itself, because you are proposing to change the basic framework of packer delivering to VM’s.
(Of course there may be  other mechanism of supporting redundancy, however it will not be as efficient as that of handing at packet level).

[cid:image009.png at 01CFF301.6EB2C440]


Thanks & regards,
Keshava

From: Alan Kavanagh [mailto:alan.kavanagh at ericsson.com]
Sent: Tuesday, October 28, 2014 6:48 PM
To: OpenStack Development Mailing List (not for usage questions)
Subject: Re: [openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

Hi Salvatore

Inline below.

From: Salvatore Orlando [mailto:sorlando at nicira.com]
Sent: October-28-14 12:37 PM
To: OpenStack Development Mailing List (not for usage questions)
Subject: Re: [openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

Keshava,

I think the thread is not going a bit off its stated topic - which is to discuss the various proposed approaches to vlan trunking.
Regarding your last post, I'm not sure I saw either spec implying that at the data plane level every instance attached to a trunk will be implemented as a different network stack.
AK--> Agree
Also, quoting the principle earlier cited in this thread -  "make the easy stuff easy and the hard stuff possible" - I would say that unless five 9s is a minimum requirement for a NFV application, we might start worrying about it once we have the bare minimum set of tools for allowing a NFV application over a neutron network.
AK--> five 9’s is a 100% must requirement for NFV, but lets ensure we don’t mix up what the underlay service needs to guarantee and what openstack needs to do to ensure this type of service. Would agree, we should focus more on having the right configuration sets for onboarding NFV which is what Openstack needs to ensure is exposed then what is used underneath guarantee the 5 9’s is a separate matter.
I think Ian has done a good job in explaining that while both approaches considered here address trunking for NFV use cases, they propose alternative implementations which can be leveraged in different way by NFV applications. I do not see now a reason for which we should not allow NFV apps to leverage a trunk network or create port-aware VLANs (or maybe you can even have VLAN aware ports which tap into a trunk network?)
AK--> Agree, I think we can hammer this out once and for all in Paris…….this feature has been lingering too long.
We may continue discussing the pros and cons of each approach - but to me it's now just a matter of choosing the best solution for exposing them at the API layer. At the control/data plane layer, it seems to me that trunk networks are pretty much straightforward. VLAN aware ports are instead a bit more convoluted, but not excessively complicated in my opinion.
AK--> My thinking too Salvatore, lets ensure the right elements are exposed at API Layer, I would also go a little further to ensure we get those feature sets to be supported into the Core API (another can of worms discussion but we need to have it).
Salvatore


On 28 October 2014 11:55, A, Keshava <keshava.a at hp.com<mailto:keshava.a at hp.com>> wrote:
Hi,
Pl find my reply ..


Regards,
keshava

From: Alan Kavanagh [mailto:alan.kavanagh at ericsson.com<mailto:alan.kavanagh at ericsson.com>]
Sent: Tuesday, October 28, 2014 3:35 PM

To: OpenStack Development Mailing List (not for usage questions)
Subject: Re: [openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

Hi
Please find some additions to Ian and responses below.
/Alan

From: A, Keshava [mailto:keshava.a at hp.com]
Sent: October-28-14 9:57 AM
To: OpenStack Development Mailing List (not for usage questions)
Subject: Re: [openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

Hi,
Pl fine the reply for the same.

Regards,
keshava

From: Ian Wells [mailto:ijw.ubuntu at cack.org.uk]
Sent: Tuesday, October 28, 2014 1:11 PM
To: OpenStack Development Mailing List (not for usage questions)
Subject: Re: [openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

This all appears to be referring to trunking ports, rather than anything else, so I've addressed the points in that respect.
On 28 October 2014 00:03, A, Keshava <keshava.a at hp.com<mailto:keshava.a at hp.com>> wrote:
Hi,

1.       How many Trunk ports can be created ?
Why would there be a limit?

Will there be any Active-Standby concepts will be there ?
I don't believe active-standby, or any HA concept, is directly relevant.  Did you have something in mind?
For the NFV kind of the scenario, it is very much required to run the ‘Service VM’ in Active and Standby Mode.
AK--> We have a different view on this, the “application runs as a pair” of which the application either runs in active-active or active standby…this has nothing to do with HA, its down to the application and how its provisioned and configured via Openstack. So agree with Ian on this.
Standby is more of passive entity and will not take any action to external network. It will be passive consumer of the packet/information.
AK--> Why would we need to care?
In that scenario it will be very meaningful to have
“Active port – connected to  “Active  Service VM”.
“Standby port – connected to ‘Standby Service VM’. Which will turn Active when old Active-VM goes down  ?
AK--> Cant you just have two VM’s and then via a controller decide how to address MAC+IP_Address control…..FYI…most NFV Apps have that built-in today.
Let us know others opinion about this concept.
AK-->Perhaps I am miss reading this but I don’t understand what this would provide as opposed to having two VM’s instantiated and running, why does Neutron need to care about the port state between these two VM’s? Similarly its better to just have 2 or more VM’s up and the application will be able to address when failover occurs/requires. Lets keep it simple and not mix up with what the apps do inside the containment.

Keshava:
Since this is solution is more for Carrier Grade NFV Service VM, I have below points to make.
Let’s us say Service-VM running is BGP or BGP-VPN or ‘MPLS + LDP + BGP-VPN’.
When such kind of carrier grade service are running, how to provide the Five-9  HA ?
In my opinion,
                Both (Active,/Standby) Service-VM to hook same underlying OpenStack infrastructure stack (br-ext->br-int->qxx-> VMa)
However ‘active VM’ can hooks to  ‘active-port’  and ‘standby VM’ hook to ‘passive-port’ with in same stack.

Instead if Active and Standby VM hooks to 2 different stack (br-ext1->br-int1 -->qxx1-> VM-active) and (br-ext2->br-int2->qxx2-> VM-Standby) can those Service-VM achieve the 99.99999 reliability ?

                Yes I may be thinking little  complicated  way from open-stack perspective..

 2.       Is it possible to configure multiple IP address configured on these ports ?
Yes, in the sense that you can have addresses per port.  The usual restrictions to ports would apply, and they don't currently allow multiple IP addresses (with the exception of the address-pair extension).

In case IPv6 there can be multiple primary address configured will this be supported ?
No reason why not - we're expecting to re-use the usual port, so you'd expect the features there to apply (in addition to having multiple sets of subnet on a trunking port).

 3.       If required can these ports can be aggregated into single one dynamically ?
That's not really relevant to trunk ports or networks.

 4.       Will there be requirement to handle Nested tagged packet on such interfaces ?
For trunking ports, I don't believe anyone was considering it.





Thanks & Regards,
Keshava

From: Ian Wells [mailto:ijw.ubuntu at cack.org.uk<mailto:ijw.ubuntu at cack.org.uk>]
Sent: Monday, October 27, 2014 9:45 PM
To: OpenStack Development Mailing List (not for usage questions)
Subject: Re: [openstack-dev] [neutron] [nfv] VM-based VLAN trunking blueprints

On 25 October 2014 15:36, Erik Moe <erik.moe at ericsson.com<mailto:erik.moe at ericsson.com>> wrote:
Then I tried to just use the trunk network as a plain pipe to the L2-gateway and connect to normal Neutron networks. One issue is that the L2-gateway will bridge the networks, but the services in the network you bridge to is unaware of your existence. This IMO is ok then bridging Neutron network to some remote network, but if you have an Neutron VM and want to utilize various resources in another Neutron network (since the one you sit on does not have any resources), things gets, let’s say non streamlined.

Indeed.  However, non-streamlined is not the end of the world, and I wouldn't want to have to tag all VLANs a port is using on the port in advance of using it (this works for some use cases, and makes others difficult, particularly if you just want a native trunk and are happy for Openstack not to have insight into what's going on on the wire).

 Another issue with trunk network is that it puts new requirements on the infrastructure. It needs to be able to handle VLAN tagged frames. For a VLAN based network it would be QinQ.

Yes, and that's the point of the VLAN trunk spec, where we flag a network as passing VLAN tagged packets; if the operator-chosen network implementation doesn't support trunks, the API can refuse to make a trunk network.  Without it we're still in the situation that on some clouds passing VLANs works and on others it doesn't, and that the tenant can't actually tell in advance which sort of cloud they're working on.
Trunk networks are a requirement for some use cases independent of the port awareness of VLANs.  Based on the maxim, 'make the easy stuff easy and the hard stuff possible' we can't just say 'no Neutron network passes VLAN tagged packets'.  And even if we did, we're evading a problem that exists with exactly one sort of network infrastructure - VLAN tagging for network separation - while making it hard to use for all of the many other cases in which it would work just fine.

In summary, if we did port-based VLAN knowledge I would want to be able to use VLANs without having to use it (in much the same way that I would like, in certain circumstances, not to have to use Openstack's address allocation and DHCP - it's nice that I can, but I shouldn't be forced to).
My requirements were to have low/no extra cost for VMs using VLAN trunks compared to normal ports, no new bottlenecks/single point of failure. Due to this and previous issues I implemented the L2 gateway in a distributed fashion and since trunk network could not be realized in reality I only had them in the model and optimized them away.

Again, this is down to your choice of VLAN tagged networking and/or the OVS ML2 driver; it doesn't apply to all deployments.

But the L2-gateway + trunk network has a flexible API, what if someone connects two VMs to one trunk network, well, hard to optimize away.

That's certainly true, but it wasn't really intended to be optimised away.
Anyway, due to these and other issues, I limited my scope and switched to the current trunk port/subport model.

The code that is for review is functional, you can boot a VM with a trunk port + subports (each subport maps to a VLAN). The VM can send/receive VLAN traffic. You can add/remove subports on a running VM. You can specify IP address per subport and use DHCP to retrieve them etc.

I'm coming to realise that the two solutions address different needs - the VLAN port one is much more useful for cases where you know what's going on in the network and you want Openstack to help, but it's just not broad enough to solve every problem.  It may well be that we want both solutions, in which case we just need to agree that 'we shouldn't do trunk networking because VLAN aware ports solve this problem' is not a valid argument during spec review.
--
Ian.

_______________________________________________
OpenStack-dev mailing list
OpenStack-dev at lists.openstack.org<mailto:OpenStack-dev at lists.openstack.org>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev


_______________________________________________
OpenStack-dev mailing list
OpenStack-dev at lists.openstack.org<mailto:OpenStack-dev at lists.openstack.org>
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.emz
Type: application/octet-stream
Size: 6228 bytes
Desc: image001.emz
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0004.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image003.emz
Type: application/octet-stream
Size: 1383 bytes
Desc: image003.emz
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0005.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: oledata.mso
Type: application/octet-stream
Size: 34030 bytes
Desc: oledata.mso
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0006.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image007.png
Type: image/png
Size: 1219 bytes
Desc: image007.png
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0003.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image008.emz
Type: application/octet-stream
Size: 7513 bytes
Desc: image008.emz
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0007.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image009.png
Type: image/png
Size: 13527 bytes
Desc: image009.png
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0004.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image004.png
Type: image/png
Size: 10539 bytes
Desc: image004.png
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20141028/26782398/attachment-0005.png>


More information about the OpenStack-dev mailing list