Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Perfect. The agent will have all static hooks for the extensions in place, like they are used in todays agents (the modular agent was derived from existing lb agent). The knowledge which concrete extension implementation to chose (e.g. lb) comes from the implementation specific manager class that is required for instantiating the modular agent. So it is ensured that with lb you get the lb extensions, with sriov you get the sriov extensions. There are no plans to make extensions more "modular" (whatever this means in this context) as well in the first round. But we can discuss for a second stage. Thanks -- Andreas (IRC: scheuran) On Mi, 2015-11-18 at 15:28 +0100, Ihar Hrachyshka wrote: > Andreas Scheuring wrote: > > > Hi all, > > I wonder if this is somehow in conflict with the modular l2 agent > > approach I'm currently following up for linuxbridge, macvtap and sriov? > > - RFE: [1] > > - Frist patchset [2] > > > > I don't think so, but to be sure I wanted to raise it up. > > I don’t believe it’s in conflict, though generally, I suggest you move > extensions code into modular l2 agent pieces, if possible. We will have > extensions enabled for lb, ovs, and sr-iov the least in Mitaka. > > Ihar > > __ > OpenStack Development Mailing List (not for usage questions) > Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Andreas Scheuring wrote: Hi all, I wonder if this is somehow in conflict with the modular l2 agent approach I'm currently following up for linuxbridge, macvtap and sriov? - RFE: [1] - Frist patchset [2] I don't think so, but to be sure I wanted to raise it up. I don’t believe it’s in conflict, though generally, I suggest you move extensions code into modular l2 agent pieces, if possible. We will have extensions enabled for lb, ovs, and sr-iov the least in Mitaka. Ihar __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Hi all, I wonder if this is somehow in conflict with the modular l2 agent approach I'm currently following up for linuxbridge, macvtap and sriov? - RFE: [1] - Frist patchset [2] I don't think so, but to be sure I wanted to raise it up. [1] https://bugs.launchpad.net/neutron/+bug/1468803 [2] https://review.openstack.org/#/c/246318/ -- Andreas (IRC: scheuran) On Mo, 2015-11-16 at 20:42 +, Cathy Zhang wrote: > I have updated the etherpad to add "Overall requirement - API cross check > among the features that can manipulate the same flow's forwarding behavior". > > Thanks, > Cathy > > -Original Message- > From: Paul Carver [mailto:pcar...@paulcarver.us] > Sent: Monday, November 16, 2015 7:50 AM > To: OpenStack Development Mailing List (not for usage questions) > Subject: Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension > access agent methods ? > > On 11/13/2015 7:03 PM, Henry Fourie wrote: > > > > > I wonder whether just pushing flows into the existing tables at random > > points in time can be unstable and break the usual flow assumed by the main > > agent loop. > > LF> No not expect any issues. > > > > Am I making sense? > > > > [1] https://etherpad.openstack.org/p/l2-agent-extensions-api-expansion > > > > I attempted to describe a possible issue at the bottom of the Etherpad in the > bullet point "Overall requirement - Flow prioritization mechanism" > > > > __ > OpenStack Development Mailing List (not for usage questions) > Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev > > __ > OpenStack Development Mailing List (not for usage questions) > Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev > __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
I have updated the etherpad to add "Overall requirement - API cross check among the features that can manipulate the same flow's forwarding behavior". Thanks, Cathy -Original Message- From: Paul Carver [mailto:pcar...@paulcarver.us] Sent: Monday, November 16, 2015 7:50 AM To: OpenStack Development Mailing List (not for usage questions) Subject: Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ? On 11/13/2015 7:03 PM, Henry Fourie wrote: > > I wonder whether just pushing flows into the existing tables at random > points in time can be unstable and break the usual flow assumed by the main > agent loop. > LF> No not expect any issues. > > Am I making sense? > > [1] https://etherpad.openstack.org/p/l2-agent-extensions-api-expansion > I attempted to describe a possible issue at the bottom of the Etherpad in the bullet point "Overall requirement - Flow prioritization mechanism" __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
On 11/13/2015 7:03 PM, Henry Fourie wrote: I wonder whether just pushing flows into the existing tables at random points in time can be unstable and break the usual flow assumed by the main agent loop. LF> No not expect any issues. Am I making sense? [1] https://etherpad.openstack.org/p/l2-agent-extensions-api-expansion I attempted to describe a possible issue at the bottom of the Etherpad in the bullet point "Overall requirement - Flow prioritization mechanism" __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Ihar, See inline. - Louis -Original Message- From: Ihar Hrachyshka [mailto:ihrac...@redhat.com] Sent: Thursday, November 12, 2015 1:12 PM To: OpenStack Development Mailing List (not for usage questions) Subject: Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ? Henry Fourie wrote: > Ihar, >Networking-sfc installs flows on br-int and br-tun for steering > traffic to the SFC port-pairs. On each bridge additional tables are > used for an egress forwarding pipeline (from the service VM port) and > an ingress pipeline (to the service VM port). Rpc operations between > the OVS driver and agents is used to initiate the flow installation. > > We'd like to work with you on defining the L2 extensions. Hi Henry, thanks for taking time to specify your needs. Could you please update the etherpad [1] with these details? LF> Will do. Speaking of new ovs tables you need, how do you reference to them? LF> br-int has one new table SF_SELECTOR in the ingress pipeline that is used steer traffic to the correct service VM by matching on the NSP (Network Service Path) of the MPLS (NSH) header. br-tun has two new tables in the egress pipeline: GRP_SELECTOR used to select a Group Table by matching on the NSP of the MPLS (NSH) header and a set of Group Tables to perform load distribution for a port-pair group. Is there any ordering guarantees for flows you will need to set that should be provided in scope of that public API of the agent? LF> No. I wonder whether just pushing flows into the existing tables at random points in time can be unstable and break the usual flow assumed by the main agent loop. LF> No not expect any issues. Am I making sense? [1] https://etherpad.openstack.org/p/l2-agent-extensions-api-expansion Ihar __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Henry Fourie wrote: Ihar, Networking-sfc installs flows on br-int and br-tun for steering traffic to the SFC port-pairs. On each bridge additional tables are used for an egress forwarding pipeline (from the service VM port) and an ingress pipeline (to the service VM port). Rpc operations between the OVS driver and agents is used to initiate the flow installation. We'd like to work with you on defining the L2 extensions. Hi Henry, thanks for taking time to specify your needs. Could you please update the etherpad [1] with these details? Speaking of new ovs tables you need, how do you reference to them? Is there any ordering guarantees for flows you will need to set that should be provided in scope of that public API of the agent? I wonder whether just pushing flows into the existing tables at random points in time can be unstable and break the usual flow assumed by the main agent loop. Am I making sense? [1] https://etherpad.openstack.org/p/l2-agent-extensions-api-expansion Ihar __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
On 11/9/2015 9:59 PM, Vikram Choudhary wrote: Hi Cathy, Could you please check on this. My mother passed away yesterday and I will be on leave for couple of weeks. I'm very sorry to hear that. Please take all the time you need. __ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Ihar, Networking-sfc installs flows on br-int and br-tun for steering traffic to the SFC port-pairs. On each bridge additional tables are used for an egress forwarding pipeline (from the service VM port) and an ingress pipeline (to the service VM port). Rpc operations between the OVS driver and agents is used to initiate the flow installation. We'd like to work with you on defining the L2 extensions. - Louis - -Original Message- From: Ihar Hrachyshka [mailto:ihrac...@redhat.com] Sent: Monday, November 09, 2015 4:44 AM To: OpenStack Development Mailing List (not for usage questions) Subject: Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ? Thanks Thomas, much appreciated. I need to admit that we haven’t heard from SFC folks just yet. I will try to raise awareness that we wait for their feedback today on team meeting. Adding [sfc] tag to the topic to get more attention. Ihar Thomas Morin wrote: > Hi Ihar, > > Ihar Hrachyshka : >> Reviving the thread. >> [...] (I appreciate if someone checks me on the following though): > > This is an excellent recap. > >> I set up a new etherpad to collect feedback from subprojects [2]. > > I've filled in details for networking-bgpvpn. > Please tell me if you need more information. > >> Once we collect use cases there and agree on agent API for extensions >> (even if per agent type), we will implement it and define as stable >> API, then pass objects that implement the API into extensions thru >> extension manager. If extensions support multiple agent types, they >> can still distinguish between which API to use based on agent type >> string passed into extension manager. >> >> I really hope we start to collect use cases early so that we have >> time to polish agent API and make it part of l2 extensions earlier in >> Mitaka cycle. > > We'll be happy to validate the applicability of this approach as soon > as something is ready. > > Thanks for taking up this work! > > -Thomas > > > >> Ihar Hrachyshka wrote: >> >>>> On 30 Sep 2015, at 12:53, Miguel Angel Ajo wrote: >>>> >>>> >>>> >>>> Ihar Hrachyshka wrote: >>>>>> On 30 Sep 2015, at 12:08, thomas.mo...@orange.com wrote: >>>>>> >>>>>> Hi Ihar, >>>>>> >>>>>> Ihar Hrachyshka : >>>>>>>> Miguel Angel Ajo : >>>>>>>>> Do you have a rough idea of what operations you may need to do? >>>>>>>> Right now, what bagpipe driver for networking-bgpvpn needs to >>>>>>>> interact with is: >>>>>>>> - int_br OVSBridge (read-only) >>>>>>>> - tun_br OVSBridge (add patch port, add flows) >>>>>>>> - patch_int_ofport port number (read-only) >>>>>>>> - local_vlan_map dict (read-only) >>>>>>>> - setup_entry_for_arp_reply method (called to add static ARP >>>>>>>> entries) >>>>>>> Sounds very tightly coupled to OVS agent. >>>>>>>>> Please bear in mind, the extension interface will be available >>>>>>>>> from different agent types (OVS, SR-IOV, [eventually LB]), so >>>>>>>>> this interface you're talking about could also serve as a >>>>>>>>> translation driver for the agents (where the translation is >>>>>>>>> possible), I totally understand that most extensions are >>>>>>>>> specific agent bound, and we must be able to identify the >>>>>>>>> agent we're serving back exactly. >>>>>>>> Yes, I do have this in mind, but what we've identified for now >>>>>>>> seems to be OVS specific. >>>>>>> Indeed it does. Maybe you can try to define the needed pieces in >>>>>>> high level actions, not internal objects you need to access to. >>>>>>> Like ‘- connect endpoint X to Y’, ‘determine segmentation id for >>>>>>> a network’ etc. >>>>>> I've been thinking about this, but would tend to reach the >>>>>> conclusion that the things we need to interact with are pretty >>>>>> hard to abstract out into something that would be generic across >>>>>> different agents. Everything we need to do in our case relates >>>>>> to how the agents use bridges and represent networks internally: &g
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Hi Vikram, Thanks for the heads-up. Take care of yourself and your family. We will provide the feedback on L2 agent. Thanks, Cathy From: Vikram Choudhary [mailto:viks...@gmail.com] Sent: Monday, November 09, 2015 6:59 PM To: OpenStack Development Mailing List (not for usage questions); Cathy Zhang Subject: Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ? Hi Cathy, Could you please check on this. My mother passed away yesterday and I will be on leave for couple of weeks. Thanks Vikram On 09-Nov-2015 6:15 pm, "Ihar Hrachyshka" mailto:ihrac...@redhat.com>> wrote: Thanks Thomas, much appreciated. I need to admit that we haven’t heard from SFC folks just yet. I will try to raise awareness that we wait for their feedback today on team meeting. Adding [sfc] tag to the topic to get more attention. Ihar Thomas Morin mailto:thomas.mo...@orange.com>> wrote: Hi Ihar, Ihar Hrachyshka : Reviving the thread. [...] (I appreciate if someone checks me on the following though): This is an excellent recap. I set up a new etherpad to collect feedback from subprojects [2]. I've filled in details for networking-bgpvpn. Please tell me if you need more information. Once we collect use cases there and agree on agent API for extensions (even if per agent type), we will implement it and define as stable API, then pass objects that implement the API into extensions thru extension manager. If extensions support multiple agent types, they can still distinguish between which API to use based on agent type string passed into extension manager. I really hope we start to collect use cases early so that we have time to polish agent API and make it part of l2 extensions earlier in Mitaka cycle. We'll be happy to validate the applicability of this approach as soon as something is ready. Thanks for taking up this work! -Thomas Ihar Hrachyshka mailto:ihrac...@redhat.com>> wrote: On 30 Sep 2015, at 12:53, Miguel Angel Ajo mailto:mangel...@redhat.com>> wrote: Ihar Hrachyshka wrote: On 30 Sep 2015, at 12:08, thomas.mo...@orange.com<mailto:thomas.mo...@orange.com> wrote: Hi Ihar, Ihar Hrachyshka : Miguel Angel Ajo : Do you have a rough idea of what operations you may need to do? Right now, what bagpipe driver for networking-bgpvpn needs to interact with is: - int_br OVSBridge (read-only) - tun_br OVSBridge (add patch port, add flows) - patch_int_ofport port number (read-only) - local_vlan_map dict (read-only) - setup_entry_for_arp_reply method (called to add static ARP entries) Sounds very tightly coupled to OVS agent. Please bear in mind, the extension interface will be available from different agent types (OVS, SR-IOV, [eventually LB]), so this interface you're talking about could also serve as a translation driver for the agents (where the translation is possible), I totally understand that most extensions are specific agent bound, and we must be able to identify the agent we're serving back exactly. Yes, I do have this in mind, but what we've identified for now seems to be OVS specific. Indeed it does. Maybe you can try to define the needed pieces in high level actions, not internal objects you need to access to. Like ‘- connect endpoint X to Y’, ‘determine segmentation id for a network’ etc. I've been thinking about this, but would tend to reach the conclusion that the things we need to interact with are pretty hard to abstract out into something that would be generic across different agents. Everything we need to do in our case relates to how the agents use bridges and represent networks internally: linuxbridge has one bridge per Network, while OVS has a limited number of bridges playing different roles for all networks with internal segmentation. To look at the two things you mention: - "connect endpoint X to Y" : what we need to do is redirect the traffic destinated to the gateway of a Neutron network, to the thing that will do the MPLS forwarding for the right BGP VPN context (called VRF), in our case br-mpls (that could be done with an OVS table too) ; that action might be abstracted out to hide the details specific to OVS, but I'm not sure on how to name the destination in a way that would be agnostic to these details, and this is not really relevant to do until we have a relevant context in which the linuxbridge would pass packets to something doing MPLS forwarding (OVS is currently the only option we support for MPLS forwarding, and it does not really make sense to mix linuxbridge for Neutron L2/L3 and OVS for MPLS) - "determine segmentation id for a network": this is something really OVS-agent-specific, the linuxbridge agent uses multiple linux bridges, and does not rely on internal segmentation Completely abstracting out packet forwarding pipelines in OVS and linuxbridge agents would possibly allow defining an interface that agen
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Hi Ihar, Thanks for initiating this discussion. I am in OPNFV Summit. Henry Fourie of SFC project team will reply with our feedback. Cathy -Original Message- From: Ihar Hrachyshka [mailto:ihrac...@redhat.com] Sent: Monday, November 09, 2015 4:44 AM To: OpenStack Development Mailing List (not for usage questions) Subject: Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ? Thanks Thomas, much appreciated. I need to admit that we haven’t heard from SFC folks just yet. I will try to raise awareness that we wait for their feedback today on team meeting. Adding [sfc] tag to the topic to get more attention. Ihar Thomas Morin wrote: > Hi Ihar, > > Ihar Hrachyshka : >> Reviving the thread. >> [...] (I appreciate if someone checks me on the following though): > > This is an excellent recap. > >> I set up a new etherpad to collect feedback from subprojects [2]. > > I've filled in details for networking-bgpvpn. > Please tell me if you need more information. > >> Once we collect use cases there and agree on agent API for extensions >> (even if per agent type), we will implement it and define as stable >> API, then pass objects that implement the API into extensions thru >> extension manager. If extensions support multiple agent types, they >> can still distinguish between which API to use based on agent type >> string passed into extension manager. >> >> I really hope we start to collect use cases early so that we have >> time to polish agent API and make it part of l2 extensions earlier in >> Mitaka cycle. > > We'll be happy to validate the applicability of this approach as soon > as something is ready. > > Thanks for taking up this work! > > -Thomas > > > >> Ihar Hrachyshka wrote: >> >>>> On 30 Sep 2015, at 12:53, Miguel Angel Ajo wrote: >>>> >>>> >>>> >>>> Ihar Hrachyshka wrote: >>>>>> On 30 Sep 2015, at 12:08, thomas.mo...@orange.com wrote: >>>>>> >>>>>> Hi Ihar, >>>>>> >>>>>> Ihar Hrachyshka : >>>>>>>> Miguel Angel Ajo : >>>>>>>>> Do you have a rough idea of what operations you may need to do? >>>>>>>> Right now, what bagpipe driver for networking-bgpvpn needs to >>>>>>>> interact with is: >>>>>>>> - int_br OVSBridge (read-only) >>>>>>>> - tun_br OVSBridge (add patch port, add flows) >>>>>>>> - patch_int_ofport port number (read-only) >>>>>>>> - local_vlan_map dict (read-only) >>>>>>>> - setup_entry_for_arp_reply method (called to add static ARP >>>>>>>> entries) >>>>>>> Sounds very tightly coupled to OVS agent. >>>>>>>>> Please bear in mind, the extension interface will be available >>>>>>>>> from different agent types (OVS, SR-IOV, [eventually LB]), so >>>>>>>>> this interface you're talking about could also serve as a >>>>>>>>> translation driver for the agents (where the translation is >>>>>>>>> possible), I totally understand that most extensions are >>>>>>>>> specific agent bound, and we must be able to identify the >>>>>>>>> agent we're serving back exactly. >>>>>>>> Yes, I do have this in mind, but what we've identified for now >>>>>>>> seems to be OVS specific. >>>>>>> Indeed it does. Maybe you can try to define the needed pieces in >>>>>>> high level actions, not internal objects you need to access to. >>>>>>> Like ‘- connect endpoint X to Y’, ‘determine segmentation id for >>>>>>> a network’ etc. >>>>>> I've been thinking about this, but would tend to reach the >>>>>> conclusion that the things we need to interact with are pretty >>>>>> hard to abstract out into something that would be generic across >>>>>> different agents. Everything we need to do in our case relates >>>>>> to how the agents use bridges and represent networks internally: >>>>>> linuxbridge has one bridge per Network, while OVS has a limited >>>>>> number of bridges playing different roles for all networks with >>>>>> internal segmentation. >>>>>> >>>>>> To loo
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Hi Cathy, Could you please check on this. My mother passed away yesterday and I will be on leave for couple of weeks. Thanks Vikram On 09-Nov-2015 6:15 pm, "Ihar Hrachyshka" wrote: > Thanks Thomas, much appreciated. > > I need to admit that we haven’t heard from SFC folks just yet. I will try > to raise awareness that we wait for their feedback today on team meeting. > Adding [sfc] tag to the topic to get more attention. > > Ihar > > Thomas Morin wrote: > > Hi Ihar, >> >> Ihar Hrachyshka : >> >>> Reviving the thread. >>> [...] (I appreciate if someone checks me on the following though): >>> >> >> This is an excellent recap. >> >> I set up a new etherpad to collect feedback from subprojects [2]. >>> >> >> I've filled in details for networking-bgpvpn. >> Please tell me if you need more information. >> >> Once we collect use cases there and agree on agent API for extensions >>> (even if per agent type), we will implement it and define as stable API, >>> then pass objects that implement the API into extensions thru extension >>> manager. If extensions support multiple agent types, they can still >>> distinguish between which API to use based on agent type string passed into >>> extension manager. >>> >>> I really hope we start to collect use cases early so that we have time >>> to polish agent API and make it part of l2 extensions earlier in Mitaka >>> cycle. >>> >> >> We'll be happy to validate the applicability of this approach as soon as >> something is ready. >> >> Thanks for taking up this work! >> >> -Thomas >> >> >> >> Ihar Hrachyshka wrote: >>> >>> On 30 Sep 2015, at 12:53, Miguel Angel Ajo wrote: > > > > Ihar Hrachyshka wrote: > >> On 30 Sep 2015, at 12:08, thomas.mo...@orange.com wrote: >>> >>> Hi Ihar, >>> >>> Ihar Hrachyshka : >>> Miguel Angel Ajo : > >> Do you have a rough idea of what operations you may need to do? >> > Right now, what bagpipe driver for networking-bgpvpn needs to > interact with is: > - int_br OVSBridge (read-only) > - tun_br OVSBridge (add patch port, add flows) > - patch_int_ofport port number (read-only) > - local_vlan_map dict (read-only) > - setup_entry_for_arp_reply method (called to add static ARP > entries) > Sounds very tightly coupled to OVS agent. > Please bear in mind, the extension interface will be available >> from different agent types >> (OVS, SR-IOV, [eventually LB]), so this interface you're talking >> about could also serve as >> a translation driver for the agents (where the translation is >> possible), I totally understand >> that most extensions are specific agent bound, and we must be >> able to identify >> the agent we're serving back exactly. >> > Yes, I do have this in mind, but what we've identified for now > seems to be OVS specific. > Indeed it does. Maybe you can try to define the needed pieces in high level actions, not internal objects you need to access to. Like ‘- connect endpoint X to Y’, ‘determine segmentation id for a network’ etc. >>> I've been thinking about this, but would tend to reach the >>> conclusion that the things we need to interact with are pretty hard to >>> abstract out into something that would be generic across different >>> agents. >>> Everything we need to do in our case relates to how the agents use >>> bridges >>> and represent networks internally: linuxbridge has one bridge per >>> Network, >>> while OVS has a limited number of bridges playing different roles for >>> all >>> networks with internal segmentation. >>> >>> To look at the two things you mention: >>> - "connect endpoint X to Y" : what we need to do is redirect the >>> traffic destinated to the gateway of a Neutron network, to the thing >>> that >>> will do the MPLS forwarding for the right BGP VPN context (called VRF), >>> in >>> our case br-mpls (that could be done with an OVS table too) ; that >>> action >>> might be abstracted out to hide the details specific to OVS, but I'm not >>> sure on how to name the destination in a way that would be agnostic to >>> these details, and this is not really relevant to do until we have a >>> relevant context in which the linuxbridge would pass packets to >>> something >>> doing MPLS forwarding (OVS is currently the only option we support for >>> MPLS >>> forwarding, and it does not really make sense to mix linuxbridge for >>> Neutron L2/L3 and OVS for MPLS) >>> - "determine segmentation id for a network": this is something >>> really OVS-agent-specific, the linuxbridge agent uses multiple linux >>> bridges, and does not rely on internal segmentation >>> >>>
Re: [openstack-dev] [neutron][sfc] How could an L2 agent extension access agent methods ?
Thanks Thomas, much appreciated. I need to admit that we haven’t heard from SFC folks just yet. I will try to raise awareness that we wait for their feedback today on team meeting. Adding [sfc] tag to the topic to get more attention. Ihar Thomas Morin wrote: Hi Ihar, Ihar Hrachyshka : Reviving the thread. [...] (I appreciate if someone checks me on the following though): This is an excellent recap. I set up a new etherpad to collect feedback from subprojects [2]. I've filled in details for networking-bgpvpn. Please tell me if you need more information. Once we collect use cases there and agree on agent API for extensions (even if per agent type), we will implement it and define as stable API, then pass objects that implement the API into extensions thru extension manager. If extensions support multiple agent types, they can still distinguish between which API to use based on agent type string passed into extension manager. I really hope we start to collect use cases early so that we have time to polish agent API and make it part of l2 extensions earlier in Mitaka cycle. We'll be happy to validate the applicability of this approach as soon as something is ready. Thanks for taking up this work! -Thomas Ihar Hrachyshka wrote: On 30 Sep 2015, at 12:53, Miguel Angel Ajo wrote: Ihar Hrachyshka wrote: On 30 Sep 2015, at 12:08, thomas.mo...@orange.com wrote: Hi Ihar, Ihar Hrachyshka : Miguel Angel Ajo : Do you have a rough idea of what operations you may need to do? Right now, what bagpipe driver for networking-bgpvpn needs to interact with is: - int_br OVSBridge (read-only) - tun_br OVSBridge (add patch port, add flows) - patch_int_ofport port number (read-only) - local_vlan_map dict (read-only) - setup_entry_for_arp_reply method (called to add static ARP entries) Sounds very tightly coupled to OVS agent. Please bear in mind, the extension interface will be available from different agent types (OVS, SR-IOV, [eventually LB]), so this interface you're talking about could also serve as a translation driver for the agents (where the translation is possible), I totally understand that most extensions are specific agent bound, and we must be able to identify the agent we're serving back exactly. Yes, I do have this in mind, but what we've identified for now seems to be OVS specific. Indeed it does. Maybe you can try to define the needed pieces in high level actions, not internal objects you need to access to. Like ‘- connect endpoint X to Y’, ‘determine segmentation id for a network’ etc. I've been thinking about this, but would tend to reach the conclusion that the things we need to interact with are pretty hard to abstract out into something that would be generic across different agents. Everything we need to do in our case relates to how the agents use bridges and represent networks internally: linuxbridge has one bridge per Network, while OVS has a limited number of bridges playing different roles for all networks with internal segmentation. To look at the two things you mention: - "connect endpoint X to Y" : what we need to do is redirect the traffic destinated to the gateway of a Neutron network, to the thing that will do the MPLS forwarding for the right BGP VPN context (called VRF), in our case br-mpls (that could be done with an OVS table too) ; that action might be abstracted out to hide the details specific to OVS, but I'm not sure on how to name the destination in a way that would be agnostic to these details, and this is not really relevant to do until we have a relevant context in which the linuxbridge would pass packets to something doing MPLS forwarding (OVS is currently the only option we support for MPLS forwarding, and it does not really make sense to mix linuxbridge for Neutron L2/L3 and OVS for MPLS) - "determine segmentation id for a network": this is something really OVS-agent-specific, the linuxbridge agent uses multiple linux bridges, and does not rely on internal segmentation Completely abstracting out packet forwarding pipelines in OVS and linuxbridge agents would possibly allow defining an interface that agent extension could use without to know about anything specific to OVS or the linuxbridge, but I believe this is a very significant taks to tackle. If you look for a clean way to integrate with reference agents, then it’s something that we should try to achieve. I agree it’s not an easy thing. Just an idea: can we have a resource for traffic forwarding, similar to security groups? I know folks are not ok with extending security groups API due to compatibility reasons, so maybe fwaas is the place to experiment with it. Hopefully it will be acceptable to create an interface, even it exposes a set of methods specific to the linuxbridge agent and a set of methods specific to the OVS agent. That would mean that the agent extension that