16:18:25 <abhijitkumbhare> #startmeeting OFP IRC meeting May 21 16:18:25 <odl_meetbot> Meeting started Thu May 21 16:18:25 2015 UTC. The chair is abhijitkumbhare. Information about MeetBot at http://ci.openstack.org/meetbot.html. 16:18:25 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:18:25 <odl_meetbot> The meeting name has been set to 'ofp_irc_meeting_may_21' 16:18:38 <tbachman> #info tbachman 16:18:40 <abhijitkumbhare> IRC only - don't have a webex 16:18:58 <ttkacik> I could provide webex if needed 16:19:04 <abhijitkumbhare> Sorry my IRC seemed to show only half of you 16:19:10 <abhijitkumbhare> we can use a webex 16:19:30 <abhijitkumbhare> can you share a webex ttkacik 16:19:46 <ttkacik> give sec 16:19:50 <abhijitkumbhare> thx 16:20:56 <ttkacik> https://cisco.webex.com/join/ttkacik 16:21:00 <abhijitkumbhare> OK 16:23:03 <abhijitkumbhare> webex is taking a long time for me to join 16:23:24 <kot-begemot> This is a slightly different version, it will be pulling a few new jars 16:23:33 <ttkacik> we are waiting for you 16:26:55 <kot-begemot> FYI: the meeting moved to webex https://cisco.webex.com/join/ttkacik 16:27:20 <colindixon> kot-begemot: I’m assuming nobody is taking notes in IRC then? 16:28:23 <kot-begemot> A few more slots on the webex, please join 16:28:28 <abhijitkumbhare> Will be taking IRC notes 16:28:36 <abhijitkumbhare> But I will add a few chairs 16:29:01 <abhijitkumbhare> #chair tbachman michal_rehak colindixon 16:29:01 <odl_meetbot> Current chairs: abhijitkumbhare colindixon michal_rehak tbachman 16:29:07 <tbachman> :) 16:29:27 <abhijitkumbhare> #topic Decision regarding He / Li 16:31:48 <tbachman> #info michal_rehak says throughput test mode has lower performance than Helium code base, but the new design has many other better properties 16:32:06 <tbachman> #info michal_rehak says the latency numbers are better, and statistics collection from architectural perspective look better 16:32:18 <tbachman> #info michal_rehak asks for arguments for going with Helium vs. Lithium 16:32:36 <tbachman> #info kot-begemot says the hashing codes are a big issue — thinks code shouldn’t be frozen in this state 16:33:12 <tbachman> #info vishnoianiol asks jamoluhrsen and LuisGomez from an integration test perspective how things look 16:33:26 <tbachman> #info jamoluhrsen says he doesn’t like the way it looks with the Lithium code base — too many failures 16:33:51 <tbachman> #info rovarga says the issue here is the version of OVS (2.0) being used in integration — has too many issues, and should be using 2.3.0+ 16:34:17 <tbachman> #info rovarga says the new plugin is good a flushing out the errors in OVS that weren’t detected by the old ones 16:34:38 <rovarga> #info my understanding is that this is related to group/meters, I think 16:34:45 <tbachman> #info jamoluhrsen says if that’s the case, then we need to get the CI environment updated ASAP 16:34:54 <colindixon> so, “doesn’t work with OVS 2.0” sounds like it’s a bug not a feature 16:34:54 <tbachman> anyone have a link? 16:34:59 <tbachman> rovarga: thx! 16:35:26 <jamoluhrsen> if the explanation is correct, it makes sense. 16:35:30 <tbachman> #info ttkacik says the old plugin has issues with statistics and working with groups 16:36:00 <tbachman> #info rovarga says there is an issue asking for an upgrade of OVS from Peter 16:36:49 <tbachman> #info jamoluhrsen asks if with the Helium code base, the plugin is incorrectly dealing with a broken vswitch, which is why our tests are passing 16:37:19 <tbachman> #info jmedved asks if jamoluhrsen can verify the behavior with the newer version of OVS (2.3) 16:37:32 <tbachman> #action jamoluhrsen to verify new plugin with OVS 2.3 16:37:43 * tbachman missed vishnoianil’s question 16:38:25 <tbachman> #info kot-begemot says there is an issue between He and Li, the original match converter was flattened and unrolled. In Lithium, it’s all flat — long sequence, so patches won’t apply cleanly across both 16:39:08 <tbachman> #info kot-begemot says the only places where the prefixes have effect in Helium are the match conversion and action conversion; In Lithium, it also affects the hashing code. 16:39:20 <tbachman> #info rovarga asks if this affects IPv6 compliance, or is it a separate issue 16:39:37 <tbachman> #info kot-begemot says that the moment you apply the v6 conversion to Lithium, you have to fix the hashing code too 16:40:18 <tbachman> #info rovarga says essentially that this means the IPv6 hashing code needs to be fixed in order to correctly hash the normalized IPv6 addresses 16:40:28 <tbachman> exponential backoff ;) 16:40:55 * tbachman missed that last one 16:41:19 <colindixon> yes, it’s related to the IPv6 functionality, it means that equality will not work 16:41:29 <tbachman> #info rovarga says this seems like two interspersed problems, and instead of tackling each one individually, we’re conflating the two 16:41:45 <tbachman> #info kot-begemot says that with the current design of the hashing code, it’s difficult to do a fix for v6 16:42:21 <tbachman> #info kot-begemot says the way it’s done currently results in a very bad hash, which creates more work and is not reusable if we fix the hashing computation 16:42:55 <tbachman> #info rovarga says if we take the entire hashing and just remove it, it’s not strictly required for the Lithium code base 16:43:00 <tbachman> #info rovarga asks if that solves the IPv6 issues 16:43:18 <tbachman> #info rovarga says we’re discussing two things — normalization of IPv6 and a hashing function in one of the code bases 16:43:43 <tbachman> #info vishnoianil says there are two separate issues, but they have overlapping boundaries 16:44:17 <tbachman> #info kot-begemot says it’s a side-effect of how v6 addresses are represented; once you normalize them and apply compression correctly, hash would result in the same value across a large set of them 16:44:32 <tbachman> #info rovarga asks if that means the code will not work correctly? 16:44:50 <tbachman> #info kot-begemot says hash means at least not putting everything in the same hash bucket — becomes a performance bottleneck 16:45:04 <tbachman> #info rovarga asks if it affects correctness or not 16:45:16 <tbachman> #info kot-begemot says if we take it out and don’t use it, then most of the fixes are done 16:45:31 * tbachman hopes he’s representing this correctly 16:45:44 <tbachman> #info rovarga asks if solving this problem allows us to cut the branch and move forward 16:46:09 <tbachman> #info kot-begemot says if we take out the hash code and acceleration benefit from it, what’s the difference between the Helium and Lithium branches — what’s the benefit 16:46:22 <tbachman> #info rovarga asks for design documents to look at to compare 16:46:36 <michal_rehak> https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Lithium_Design_Proposal 16:46:43 <tbachman> #info rovarga says if you look at the design differences, the benefits should be self-evident 16:46:44 <abhijitkumbhare> #link https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Lithium_Design_Proposal 16:46:48 <tbachman> #lundo 16:46:52 <abhijitkumbhare> #link https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Lithium_Design_Proposal 16:47:01 <tbachman> #undo 16:47:01 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Link object at 0x1b886d0> 16:47:02 <tbachman> #undo 16:47:02 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Link object at 0x1b88990> 16:47:14 <tbachman> #link https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Lithium_Design_Proposal Link to design of new openflowplugin for Lithium 16:47:21 <tbachman> abhijitkumbhare: sorry — wanted to have context 16:47:29 <abhijitkumbhare> np 16:47:50 <tbachman> #info kot-begemot says he read something about changing the way the I/O works — wants to understand what has changed 16:47:56 <colindixon> frustratingly, that isn’t linked from the release plan 16:47:59 <colindixon> :-( 16:48:02 <colindixon> https://wiki.opendaylight.org/view/Openflow_Protocol_Library:Lithium_Release_Plan 16:48:21 <tbachman> #link https://wiki.opendaylight.org/view/Openflow_Protocol_Library:Lithium_Release_Plan OpenFlow plugin Lithium release plan 16:48:24 <tbachman> colindixon: thx! 16:49:12 <abhijitkumbhare> colindixon tbachman - the right link is https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Lithium_Release_Plan - the redesign is linked from there 16:49:18 <tbachman> #info vishnoianil says that as someone involved for a long time, the pros for the old code are that it’s stable and tested; the pros for the second are the better architecture, but the cons are the performance and stability - but the priority here is cutting the branch 16:49:21 <tbachman> abhijitkumbhare: thx! 16:49:41 <colindixon> ah 16:49:41 <tbachman> (and thx to colindixon for also providing that link ;) ) 16:49:49 <colindixon> it’s not linked as a wiki link 16:49:50 <colindixon> *sigh* 16:49:51 <colindixon> ok 16:50:02 <tbachman> #undo 16:50:02 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Info object at 0x192eb50> 16:50:03 <colindixon> hence this: https://wiki.opendaylight.org/view/Special:WhatLinksHere/OpenDaylight_OpenFlow_Plugin:Lithium_Design_Proposal 16:50:05 <tbachman> #undo 16:50:05 <odl_meetbot> Removing item from minutes: <MeetBot.ircmeeting.items.Link object at 0x1a27710> 16:50:29 <colindixon> so, can we make it so that users can pick which OF plugin implementation they can use 16:50:33 <colindixon> so that some people can use the newer one 16:50:42 <tbachman> #link https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Lithium_Release_Plan The OpenFlow plugin Lithium release plan, with link to redesign 16:50:53 <rovarga> colindixon: it is already thath way, there are two features 16:50:55 <tbachman> #info colindixon asks if we can make it so that users can pick which OF plugin implementation they can use 16:50:59 <colindixon> ok 16:51:06 <tbachman> #info rovarga says it’s already that way — there are two features 16:51:33 * tbachman was too busy finding links to catch what vishnoianil was saying 16:51:39 <tbachman> #info rovarga asks what’s blocking us from creating the branch 16:51:53 <tbachman> #info vishnoianil thinks the main reason was the IPv6 was the main thing blocking the branch 16:52:00 <tbachman> #info rovarga asks who owns the IPv6 issue 16:52:16 <tbachman> #info kot-begemot says that in the current state, the Lithium gerit will fix everything but the hashing code 16:52:47 <tbachman> #info rovarga asks if we take out the hashing code with those fixes, then we’re good to cut the branch 16:53:02 <abhijitkumbhare> #link https://git.opendaylight.org/gerrit/#/c/20704 Lithium gerritt 16:53:04 <tbachman> #info kot-begemot says taking it out and fixing it is not suitable to be done for freezing 16:53:21 <tbachman> #info kot-begemot says we still have this failing in CI — from a code perspective, the answer is yes 16:57:11 <tbachman> #info jmedved asks how we can solve the problem of cutting the branch 16:57:53 <tbachman> #info rovarga says we can also take this decision to the TSC 16:59:11 <tbachman> #info kot-begemot says the key question is to be whether we should use the Helium or Lithium plugin 16:59:30 <tbachman> #info rovarga says colindixon pointed out that we can ship both and let users decide (currently both features are there) 16:59:45 <tbachman> #info rovarga says what we need to solve is the problem of the stable/lithium branch 16:59:57 <tbachman> #info vishnoianil says he votes for that approach (two features) 17:00:09 <tbachman> #info abhijitkumbhare asks if we’re okay with two implementations or single implementnation 17:00:24 <tbachman> #info abhijitkumbhare says it sounds like most folks are okay with both implementations 17:00:35 <tbachman> #info abhijitkumbhare asks if anyone is opposed to both implementations 17:00:50 <tbachman> #info vishnoianil says he approves, with the default being the Helium plugin 17:01:15 <tbachman> #info abhijitkumbhare says we can make the call as to which is the default at RC0, 1, 2.... 17:01:34 <tbachman> #info kot-begemot says he’s fine with that — as long as users have a stable version to work with, he’s fine with that 17:01:57 <tbachman> #agreed: openflowplugin to provide both Helium and Lithium implementations as part of the Lithium release 17:03:19 <tbachman> #startvote Should the committers decide to cut stable/lithium +1,0,-1 17:03:19 <odl_meetbot> Unable to parse vote topic and options. 17:03:23 <tbachman> :( 17:03:38 <tbachman> #startvote Should the committers decide to cut stable/lithium? +1,0,-1 17:03:38 <odl_meetbot> Begin voting on: Should the committers decide to cut stable/lithium? Valid vote options are +1, 0, -1. 17:03:38 <odl_meetbot> Vote using '#vote OPTION'. Only your last vote counts. 17:03:51 <mbobak> #vote +1 17:04:03 <v_demcak> #vote +1 17:04:10 <michal_rehak> #vote +1 17:04:15 <abhijitkumbhare> #vote +1 17:04:20 <tbachman> abhijitkumbhare: can you do a +1 for anil? 17:04:53 <tbachman> abhijitkumbhare: should I end this vote and start another? 17:05:01 <tbachman> #endvote 17:05:01 <odl_meetbot> Voted on "Should the committers decide to cut stable/lithium?" Results are 17:05:01 <odl_meetbot> +1 (4): mbobak, v_demcak, michal_rehak, abhijitkumbhare 17:05:57 <vishnoianil> #info My vote is +1, we should go ahead with branch cutting given that Ipv?prefix issue is fixed in He code base 17:06:15 <vishnoianil> #we will still stick to both the implementation (He (default) and Li) 17:06:25 <kot-begemot> I will fix it in the Li code base, 17:06:29 <kot-begemot> without the hashing code 17:06:38 <kot-begemot> That needs a full rewrite to be fixable 17:06:48 <vishnoianil> #info we will take on more call on Rc0 about the concerns discussion by the members, on switching to new Li implementaiton 17:06:51 <tbachman> #info we will still stick to both the implementations (Helium default and Lithium) 17:07:04 <kot-begemot> In fact, the current gerrit for Li is a viable fix 17:07:07 <tbachman> #info kot-begemot says he will fix it in the Lithium code base without the hashing code 17:07:25 <vishnoianil> #info if we will not be able to take clear decision on that, we can ship both the implementation in LI release and defer it to User to use either of the feature 17:08:24 <kot-begemot> #info that part is broken as it is now, and needs a rewrite to be a hash. Presently it is not. Adding all values in header fields a hash does not make 17:09:04 <abhijitkumbhare> I think this is all getting jumbled up a bit :) 17:09:30 <abhijitkumbhare> I will try to do an agreed to it 17:09:49 <abhijitkumbhare> #agreed we should go ahead with branch cutting given that Ipv?prefix issue is fixed in He code base 17:10:03 <abhijitkumbhare> #agreed we will still stick to both the implementation (He (default) and Li) 17:10:22 <abhijitkumbhare> #agreed we will take on more call on Rc0 about the concerns discussion by the members, on switching to new Li implementaiton 17:10:39 <abhijitkumbhare> #agreed if we will not be able to take clear decision on that, we can ship both the implementation in LI release and defer it to User to use either of the feature 17:11:10 <abhijitkumbhare> Should I end meeting with this? 17:11:22 <kot-begemot> I think you covered everything 17:11:28 <abhijitkumbhare> #endmeeting