19:05:12 <Madhu> #startmeeting Weekly OVSDB meeting 19:05:12 <odl_meetbot> Meeting started Tue May 20 19:05:12 2014 UTC. The chair is Madhu. Information about MeetBot at http://ci.openstack.org/meetbot.html. 19:05:12 <odl_meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:05:12 <odl_meetbot> The meeting name has been set to 'weekly_ovsdb_meeting' 19:05:28 <tbachman> #info networkstatic needs a new headset 19:05:37 <Madhu> #info webex needs to work 19:06:12 <Madhu> #vote Hangout +1 0 -1 19:06:33 <tbachman> +1 19:06:33 <Madhu> still learning ;) 19:06:41 <tbachman> #vote +1 19:06:45 <Madhu> #startvote Hangout +1,0,-1 19:06:45 <odl_meetbot> Unable to parse vote topic and options. 19:07:06 <tbachman> Needs the question mark 19:07:13 <tbachman> Like this: 19:07:19 <tbachman> Hangout? +1,0,-1 19:07:25 <Madhu> #startvote hangout ? +1,0,-1 19:07:25 <odl_meetbot> Begin voting on: hangout ? Valid vote options are +1, 0, -1. 19:07:25 <odl_meetbot> Vote using '#vote OPTION'. Only your last vote counts. 19:07:30 <Madhu> #vote +1 19:07:30 <tbachman> #vote +1 19:09:14 <Madhu> #endvote 19:09:14 <odl_meetbot> Voted on "hangout ?" Results are 19:09:14 <odl_meetbot> +1 (2): tbachman, Madhu 19:09:26 <Madhu> #info Brent is raging ;) 19:10:26 <networkstatic> Im a QA guy 19:22:45 <networkstatic> #link Floating IP Client Commands http://docs.openstack.org/user-guide/content/floating_ip_allocate.html 19:23:13 <networkstatic> did that work? 19:24:06 <networkstatic> #info Floating IP Client Commands http://docs.openstack.org/user-guide/content/floating_ip_allocate.html 19:57:46 <jimooo99> has anyone seen this errer: 19:57:46 <jimooo99> io.netty.handler.codec.TooLongFrameException: frame length exceeds 100000: 100154 - discarded 19:58:01 <jimooo99> it's coming from ovsdb 19:58:15 <jimooo99> at org.opendaylight.ovsdb.lib.jsonrpc.JsonRpcDecoder.fail(JsonRpcDecoder.java:159) [bundlefile:na] 20:27:06 <networkstatic> GiovanniMeo is bamf 21:05:22 <networkstatic> brb 22:26:25 <padma> when trying to run the mininet with ovsdb we are seeing netty exceptions is OpenDaylight controller 22:26:30 <padma> WARN i.n.channel.DefaultChannelPipeline - An exceptionCaught() event was fired, and it reached at the tail of the pipeline. It usually means the last handler in the pipeline did not handle the exception. io.netty.handler.codec.TooLongFrameException: frame length exceeds 100000: 100000 - discarded 22:26:36 <padma> has anyone seen this 22:56:48 <odp-gerritbot> Hsin-Yi Shen proposed a change to ovsdb: BUG-1056 brocastflow is not proper modified after deleting interface https://git.opendaylight.org/gerrit/7270 23:32:59 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Adjusting the ovsdb stable/hydrogen dependencies to appropriate Controller, Yangtools dependencies. https://git.opendaylight.org/gerrit/7273 00:07:18 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Fixed typo in PortHandler.java. https://git.opendaylight.org/gerrit/7280 00:09:06 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Removed convertNeutronIDToKey(String) calls. https://git.opendaylight.org/gerrit/7281 00:11:00 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: BUG-954: Add support for Neutron port/network delete events for OF1.3 https://git.opendaylight.org/gerrit/7282 00:59:16 <odp-gerritbot> A change was merged to ovsdb: Adjusting the ovsdb stable/hydrogen dependencies to appropriate Controller, Yangtools dependencies. https://git.opendaylight.org/gerrit/7273 01:04:20 <odp-gerritbot> A change was merged to ovsdb: Fixed typo in PortHandler.java. https://git.opendaylight.org/gerrit/7280 01:06:54 <odp-gerritbot> A change was merged to ovsdb: gitreview: add .gitreivew for stable/hydrogen branch https://git.opendaylight.org/gerrit/7262 01:08:47 <odp-gerritbot> A change was merged to ovsdb: Removed convertNeutronIDToKey(String) calls. https://git.opendaylight.org/gerrit/7281 01:11:40 <odp-gerritbot> A change was merged to ovsdb: BUG-954: Add support for Neutron port/network delete events for OF1.3 https://git.opendaylight.org/gerrit/7282 01:16:44 <Madhu> networkstatic: thanks boss. 01:16:54 <networkstatic> course sir 01:17:11 <Madhu> networkstatic: do u think we shud also cherry pick the IPFIX gerrits ? 01:17:26 <Madhu> we have a couple of them ... if u want we can take them over to stable/hydrogen 01:17:48 <networkstatic> i think onlu jimoo was using, whatever u think buddy. 01:18:27 <networkstatic> getting my manual QA on now, :) 01:18:54 <Madhu> oh dude. pls don't waste time in doing QA sir :) 01:19:01 <Madhu> lets put effort on IT 01:19:07 <Madhu> and let the Scripts handle it for us 01:19:40 <networkstatic> lol, i need a working build to verify stuffs tonight, its good, have ones already. 01:19:56 <Mierdin> hey guys how was today's meeting? Hated to miss but I was in my session 01:19:57 <networkstatic> btw, u needs to DL Veep 01:20:00 <networkstatic> sooo good 01:21:28 <Madhu> networkstatic: of course sir. ur review comments are important. both in gerrit and TV shows ;) 01:21:38 <networkstatic> haha 01:21:48 <Madhu> Mierdin: it was a good call. Status updates and my UT/IT talk :) 01:22:00 <networkstatic> yes, need a little bit of pleasure in our world of living on laptop 01:22:03 <Madhu> u didn't miss that talk much as I bored u the other day :) 01:22:23 <networkstatic> one of us need to blog on the diff of UT/IT/SI 01:22:54 <Madhu> do u watch the shows in ur laptop sir ? 01:22:56 <networkstatic> might throw into slides real quick while fresh 01:23:02 <networkstatic> nah airplay to tv 01:23:02 <Madhu> I have a kick-ass HTPC at home ;) 01:23:12 <Madhu> ah... fancy 01:23:15 <networkstatic> man, i stopped the htpc w/ airplay 01:23:19 <networkstatic> no maintenance 01:23:43 <networkstatic> mainly cause faster to drag to Beamer 01:23:50 <networkstatic> i share app :) 01:23:57 <Madhu> AA :) 01:24:07 <networkstatic> u have JB ATV2? 01:24:10 <networkstatic> or 3 01:24:34 <networkstatic> or even better, mac mini 01:24:40 <networkstatic> those are awesome 01:24:47 <Madhu> u know my setup :) 01:24:53 <networkstatic> i forgot! lol 01:24:54 <Madhu> i cannot afford to use mac-mini 01:24:56 <networkstatic> haha 01:25:02 <networkstatic> laptop stuffed behind tv 01:25:03 <networkstatic> lol 01:25:12 <Madhu> nahh... 01:25:16 <networkstatic> too cheap for mini :) 01:25:19 <networkstatic> i know i am 01:25:25 <Madhu> mine need PCIE slots 01:25:30 <Madhu> oh ya. that too ;) 01:25:31 <networkstatic> those things hold value on ebay like a champ 01:26:27 <networkstatic> ooh new intellij update out 01:27:02 <Madhu> am an eclipse guy :) tried intellij but... somehow it is not clicking 01:27:09 <Madhu> maybe i have to put more time into it 01:28:03 <Madhu> networkstatic: btw.. hope u r following on the email thread on the broadcast multiple output port issue 01:28:13 <Madhu> it is an issue on modify flow operation 01:28:20 <Madhu> and is currently being fixed 01:28:30 <Madhu> pls keep in mind during your manual QA :) 01:28:41 <Madhu> DHCP won't resolve until this issue is merged 01:28:55 <Madhu> networkstatic: fyi : https://git.opendaylight.org/gerrit/#/c/7099/ 01:28:57 <networkstatic> i am buddy, i saw that earlier, only thing still broken here 01:29:50 <networkstatic> im burned out. going back in the cave for a couple days. sure im wearing out my welcome of asking questions 01:30:07 <networkstatic> http://monolingual.sourceforge.net just deleted 3Gb of language packs off mac 01:30:20 <networkstatic> more room for VMs 01:33:08 <Madhu> networkstatic: take off sir. we need the crazy networkstatic not the sad one :) 01:34:13 <networkstatic> sad faces bad :) 01:42:55 <Mierdin> madhu it is not possible for you to bore me 01:43:49 <Mierdin> oh and btw just ran into edwarnicke at clus 01:43:55 <Mierdin> had a baller chat about helium 01:44:06 <Networkn3rd> introduced me as well 01:44:08 <Networkn3rd> I feel important now 01:44:10 <Networkn3rd> ish 01:44:12 <Networkn3rd> not really... 01:49:58 * edwarnicke waves :) 07:29:11 <elrarun> Hi 13:07:55 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Commit operation as per Section 5.2.7 of RFC 7047 https://git.opendaylight.org/gerrit/7297 13:07:56 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Fixed a bug in where condition usage in Update & Delete operations. Also moved Delete operation to a seperate transaction in the Test code to test multiple transactions. https://git.opendaylight.org/gerrit/7298 13:18:13 <Madhu> MierdinMobile: hello sir. guess u r busy with #CLUS. Need some help with Mutate. 13:18:19 <Madhu> dave_tucker: hey buddy. 13:18:28 <dave_tucker> hey Madhu! 13:19:27 <Madhu> the IT that we are writing for operations Insert, delete, etc. will not work until we have Mutate 13:19:39 <Madhu> once Mutate is in place, we will need better Asserts :) 13:20:07 <Madhu> regarding Mutate, I am waiting on Matt to address the comments 13:20:56 <dave_tucker> Madhu: cool 13:20:57 <Madhu> dave_tucker: just fyi. 13:21:14 <dave_tucker> also, you need to retrigger the job for 7298 in jenkins 13:22:25 <dave_tucker> code review now done ;) 13:24:09 <dave_tucker> also we need to fix this IT - https://jenkins.opendaylight.org/ovsdb/job/ovsdb-verify/org.opendaylight.ovsdb$ovsdb_library/412/testReport/org.opendaylight.ovsdb.lib.jsonrpc/TestClient/testBasicFlow/ 13:24:25 <dave_tucker> as it's a pain in the ass that builds are being marked unstable because of this 13:25:22 <dave_tucker> oh and I retriggered the build in jenkins for you while I was there 13:25:38 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Abort operation as per Section 5.2.8 of RFC 7047 https://git.opendaylight.org/gerrit/7300 13:25:59 <Madhu> thanks sir 13:26:09 <Madhu> your +1s are useful :) 13:26:17 <Madhu> will ask Brent/Ashwin to merge it 13:26:36 <Madhu> i don't want to +2 my own commits :) 13:27:18 <Madhu> dave_tucker: sure. i will take a look @ the basic flow error 13:27:37 <Madhu> brb sir 13:27:38 <dave_tucker> just need to catch the exception and use a different port 13:27:40 <dave_tucker> ;) 13:27:52 <Madhu> lets use a different port for this IT period :) 13:28:10 <dave_tucker> lol. issue is only present when 2 jobs are building at the same time 13:29:00 <Madhu> ya. i mean pick ports randomly between 6640 - 6740 ;) 13:29:11 <dave_tucker> but then we might have a coliision 13:29:18 <dave_tucker> so still need logic to handle that ;) 13:29:19 <Madhu> brb sir. 13:29:21 <dave_tucker> k 13:36:22 <dave_tucker> when you are back we should chat about my commit you have -1'd :p 13:36:49 <Madhu> am back. :) but before chatting about that, i want to quickly get your opinion on Mutate 13:37:01 <Madhu> we both commented on Matt's Mutate diffs 13:37:23 <Madhu> but since we need that to do proper IT, I am inclined on merging them first and then I will update the tests with mutate operation in place. 13:37:35 <dave_tucker> Madhu: that works for me 13:37:37 <Madhu> I don't want to get blocked by the absence of Mutate. 13:37:41 <Madhu> okay. lemme merge them now. 13:37:57 <dave_tucker> waiting for maven before I add the IT for select fyi 13:37:58 <Madhu> now... lets talk about the static table name issue. 13:38:02 <dave_tucker> yeah :p 13:38:25 <dave_tucker> so my point is, it doesn't matter.... the constants are a convenience function and you can use them, or not 13:38:38 <Madhu> my point is 13:38:46 <dave_tucker> but based on the fact that everybody will be defining the same basic set of constants, makes sense to have them in library 13:38:49 <Madhu> the moment you declare it.. it gets first class citizen 13:39:13 <Madhu> I don't want the Table names to get that privilege in the library 13:39:44 <Madhu> We had this extensive discussion and intact am opposed to even having the database name OpenVSwitch given that first class citizen role in the library 13:39:59 <dave_tucker> well I think we should add the VTEP stuff too 13:40:03 <Madhu> and i also changed that code (will get it out of my stash and push it) 13:40:20 <dave_tucker> even if it's just to a class full of constants in utils 13:40:24 <dave_tucker> or something ;) 13:40:35 <Madhu> not in library :) 13:40:40 <Madhu> lemme put it this way 13:40:41 <dave_tucker> the library itself doesn't care - it just wants a string 13:40:51 <Madhu> if u c it in RFC7047, it gets into library :) 13:41:17 <Madhu> if u don't see it in the RFC, library should not give it a preference. 13:41:30 <Madhu> ashw7n: my man. :) 13:41:35 <dave_tucker> lol 13:41:47 <Madhu> we are stalkers ;) 13:41:53 <Madhu> dave_tucker: back to the topic 13:42:08 <Madhu> we made the mistake during the Hydrogen release by mixing library and plugin 13:42:13 <dave_tucker> appreciate your view Madhu. but as a developer, I like convenience.... and it makes it hard to use the library if I have to read the OVSDB schema to get the table names ;) 13:42:35 <Madhu> dave_tucker: that problem will be solved :) once we have plugin operating on Library 13:42:43 <dave_tucker> + DRY principle. we will have multiple defintions of the same constants 13:42:57 <Madhu> no. we will not :) 13:43:01 <dave_tucker> ok cool 13:43:06 <Madhu> lemme explain it other way 13:43:12 <dave_tucker> then in that case I'll abandon and we move this to the plugin 13:43:28 <Madhu> the tests that we are adding to the library now are "temporary" 13:43:31 <Madhu> once plugin is in place 13:43:44 <Madhu> lets get crazy there :) and use these constants and intact I would recommend static typing there 13:43:56 <Madhu> lets also introduce first class tables as classes on the plugin layer 13:44:00 <dave_tucker> Madhu: I'm not keen on static typing ;) 13:44:01 <Madhu> and add all the nice test-cases there 13:44:17 <dave_tucker> but we should DEFINATELY have IT/UT in the library 13:44:18 <Madhu> library should be kept clean. IMHO 13:44:39 <dave_tucker> and that should include the tests we have today (but with more assertions) 13:44:40 <dave_tucker> ;) 13:44:42 <Madhu> the UT/IT in library should be completely RFC7047 dependent :) 13:45:00 <Madhu> dave_tucker: +10000 on assertions 13:45:02 <dave_tucker> Madhu: yessir, but you have to test against an implementation 13:45:25 <Madhu> dave_tucker: can u imagine the world am trying to try in the library ? :) 13:45:35 <Madhu> we can use list_dbs :) to get all the databases supported 13:45:48 <dave_tucker> Madhu: ok yeah that makes sense 13:45:50 <Madhu> and then we can do getSchema to get all the tables and columns in that 13:45:54 <Madhu> with these 2 info 13:46:05 <Madhu> we can test all the tables and individual columns from library 13:46:19 <Madhu> without needing to have a string like "bridge" or "Interface" defined in the Test classes :) 13:46:21 <dave_tucker> but to do CRUD, we need to understand the schema though 13:46:27 <dave_tucker> i think 13:46:36 <Madhu> yep. CRUD needs it. but that is in the Northbound layer 13:46:58 <dave_tucker> well, ok I mean Select + Insert + Update + Delete 13:47:00 <Madhu> but even there :) if we expose the getSchema and list_dbs (which we will) 13:47:13 <Madhu> even Select +insert + update + delete 13:47:20 <Madhu> can be done entirely schema independent 13:47:33 <Madhu> just by getting the info using list_dbs and getSchema :) 13:47:42 <Madhu> because everything u need to know is in there 13:47:46 <dave_tucker> yeah I guess we just need some test data that matches the expected types as per the RFC 13:47:50 <Madhu> tableName, Column Name, Column Type. etc... 13:48:23 <Madhu> so IMO, library tests should not even have the string "bridge" represented there 13:48:32 <dave_tucker> Madhu: fair enough sir 13:48:33 <Madhu> we are adding such tests now just to make progress. 13:48:46 <Madhu> once we are in stable grounds all these tests should move to plugin 13:48:51 <dave_tucker> yeah I know... your reasoning is correct, as always :p 13:48:55 <Madhu> and we will live in a pure Library world :) 13:49:00 <dave_tucker> lol 13:49:14 <dave_tucker> where everything is abstract 13:49:21 <Madhu> since Ben & Justin are happy to have this library as ref impl in OVS 13:49:30 <Madhu> its better to be abstract and scalable to any schema 13:49:37 <Madhu> and folks just pick it up and use it :) 13:50:01 <Madhu> dave_tucker: sorry for being religious on this topic. 13:50:10 <Madhu> just trying to avoid the same mistakes that we did for Hydrogen 13:50:40 <Madhu> whenever i see these static values (even in tests), i don't feel comfortable 13:50:42 <dave_tucker> Madhu: it's fine sir. I was looking at use of the library forgetting that we needed to support multiple schemas. totally makes sense now we've spoken about it 13:50:51 <Madhu> but being practical, am just trying to make sure we make progress too 13:51:00 <Madhu> excellent buddy. 13:51:37 <Madhu> intact once all the operations are done, my plan is to write the abstract and generic test-case 13:51:49 <Madhu> which will ram through all the tables/columns and operations 13:52:07 <Madhu> without the need to have any static typing or strings :) 13:52:11 <Madhu> but that is a lot of work 13:52:40 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Add OVSDB Select Operation https://git.opendaylight.org/gerrit/6861 13:52:53 <dave_tucker> thanks odp-gerritbot 13:53:02 <dave_tucker> included your fix for the where's too Madhu 13:53:09 <dave_tucker> as well as the IT 13:53:15 <Madhu> thanks sir. 13:57:32 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Unit Test Coverage for OvsdbClientImpl https://git.opendaylight.org/gerrit/7122 13:57:45 <dave_tucker> and that one was rebased and addressed sam's comments 13:58:33 <dave_tucker> omg networkstatic_ finally cloned himself - networkstatic_zZ 13:58:49 <dave_tucker> that's too much awesome for one channel 14:01:04 <Madhu> lol 14:01:07 <Madhu> networkstatic_: u there sir ? 14:04:11 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Update .gitreview to use topic/schema branch https://git.opendaylight.org/gerrit/7303 14:04:12 <Madhu> networkstatic: hey buddy 14:16:07 <dave_tucker> Madhu: ref https://git.opendaylight.org/gerrit/7122 - do u know if we could s/Integration/IT/g with the name of the base class without confusing Maven? 14:18:11 <dave_tucker> also Madhu we can't merge that until https://git.opendaylight.org/gerrit/#/c/7121/ merges 14:18:47 <dave_tucker> when writing the UT I noticed that isReady is always true... the timeout in the todo is therefore pointless 14:21:26 <dave_tucker> "return null != schema;" is always true. Because we init schema when the class is instantiated. 14:21:48 <dave_tucker> a better check is to see if the list of schemas > 0, which is what was added in 7121 14:30:25 <dave_tucker> Madhu: if you are still there sir... re-read your comments... would you be happy with this https://gist.github.com/878b3de25df043afbc38 for the timeout 14:36:35 <dave_tucker> now using isEmpty instead of size - https://gist.github.com/f2beecf462f15eaff1fb 14:41:35 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Stop isReady method always returning true https://git.opendaylight.org/gerrit/7121 14:47:36 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Stop isReady method always returning true https://git.opendaylight.org/gerrit/7121 14:49:03 <Madhu> dave_tucker: hey buddy. 7am+ is challenging :) 14:49:18 <Madhu> be back online in 1 hr. will continue reviewing / merge the code sir 14:52:56 <dave_tucker> oops. sorry bro forgot it was 7am 16:03:54 <Mierdin> Hey Madhu 16:03:59 <Mierdin> read some of the chat on my phone 16:04:02 <Mierdin> saw you guys were asking about Mutate 16:04:10 <Madhu> Mierdin: hi buddy 16:04:23 <Mierdin> I'm not holding you guys up on anything, right? I can take a break from this and make a change if you're waiting on something 16:04:26 <Madhu> yes sir. Mutate is an important operation for us to write proper IT 16:04:43 <Mierdin> agreed 16:04:50 <Madhu> it would be awesome if you can get the Mutate pushed 16:05:20 <Madhu> I was inclined to merge the code as is 16:05:34 <Madhu> but I have a few comments and wanted to talk to u first. 16:06:20 <Mierdin> I noticed you commented on the commit I assume that's what you'd like me to fix 16:06:39 <Madhu> there is 1 more thing 16:07:23 <Madhu> Mierdin: i was reviewing the value() method : count = (Integer) value; //TODO: Verify this is the right thing to do. Fairly sure from re- 16:07:23 <Madhu> reading RFC but want to make sure. 16:07:57 <Madhu> the "count" member belongs to the result object. 16:08:28 <Madhu> also. a value() method doesn't apply for mutate operation 16:09:12 <Madhu> As we discussed, it is similar to update() but carries specific mutations 16:10:16 <GiovanniMeo> ehi guys, 16:10:22 <GiovanniMeo> i need some help with ovsdb 16:10:30 <Mierdin> gotcha. So not needed? I assumed it needed to be there since the RFC says it returns a "count" object just like Update 16:10:41 <GiovanniMeo> i'm trying to rune the autorelease to release ovsdb project 16:10:43 <Madhu> GiovanniMeo: sure 16:10:45 <GiovanniMeo> but i noticed 16:10:46 <Mierdin> Madhu: And it's good to hear that you'd prefer me to bind the Operations changes with the Mutate changes. I wasn't sure if that was the right way to do it but makes sense now that they should be together, given their relationship 16:10:49 <GiovanniMeo> this dependency 16:10:52 <GiovanniMeo> org.opendaylight.integration:distributions-virtualization:zip:osgipackage:0.1.2-SNAPSHOT:provided 16:11:11 <GiovanniMeo> so this is causing a circular dependency in the release 16:11:21 <GiovanniMeo> because ovsdb need the integration 16:11:26 <GiovanniMeo> and the integration needs ovsdb 16:11:32 <Madhu> Mierdin: right. typically i prefer smaller commits. but this specific one 16:11:52 <Madhu> Mierdin is hard to catch the issue unless it is in the package of the 3 things (which includes the Test coverage) :) 16:12:09 <Madhu> GiovanniMeo: correct :) 16:12:13 <Madhu> so ? 16:12:14 <GiovanniMeo> so i wonder 16:12:34 <GiovanniMeo> what you need the distribution for 16:12:36 <GiovanniMeo> in here 16:12:37 <Mierdin> Madhu: :) test coverage was the next change that I was working on. 16:12:53 <GiovanniMeo> so we can see if we can break the circle 16:12:55 <GiovanniMeo> somehow 16:13:18 <Madhu> Mierdin: perfect. i don't want to act like an ass... but we are trying to be more organized this time around :) 16:13:38 <Madhu> GiovanniMeo: distribution here is like the same reason all other project has 16:13:49 <Mierdin> Madhu: no, not at all! It's like I said, I want to make sure that my contributions are done the right way 16:13:53 <Madhu> for testing OVSDB with their dependent bundles 16:14:20 <Madhu> Mierdin: there is nothing wrong in not adding the tests in the commit (for the master branch and Hydrogen release) :) 16:14:37 <Madhu> GiovanniMeo: this issue is NOT specific to ovsdb I believe 16:14:55 <GiovanniMeo> all the other depends on the controller distro 16:14:59 <Madhu> GiovanniMeo: all other projects does the same... they have a distribution which pulls in the dependencies and then builds on their own 16:15:04 <GiovanniMeo> not on one from integration 16:15:07 <Madhu> GiovanniMeo: fair enough :) 16:15:18 <Mierdin> Madhu: okay so I need to head over to Moscone, but can make the changes. In summary, I need to remove that method entirely, and also merge the change to Operations.java with the addition of Mutate.java. Correct? 16:15:24 <Madhu> I was lazy because we depend on controller and open flow plugin projects 16:15:27 <Mierdin> Madhu: I can follow up once out of here with proper UT/IT 16:15:31 <GiovanniMeo> so if we build the distro in here 16:15:37 <GiovanniMeo> starting from controller base 16:15:43 <GiovanniMeo> would you see any issue with that? 16:15:45 <Madhu> Mierdin: thanks 16:16:06 <GiovanniMeo> or if we create a distro 16:16:08 <Madhu> GiovanniMeo: not a problem. But we also need the openflowplugin and openflowjava pulled in as well 16:16:29 <Madhu> the trouble we are facing is the version compatibility :( 16:16:39 <GiovanniMeo> sure that means ovsdb will be released after openflowplugin and openflowjava 16:16:50 <Madhu> as u know it is all interlinked and that is the reason i made it simple by pulling in the virt edition from integration branch 16:16:51 <GiovanniMeo> just looking to break the circular dependency 16:16:55 <GiovanniMeo> i will propose a patch 16:17:02 <GiovanniMeo> based on this 16:17:04 <GiovanniMeo> thx Madhu 16:17:50 <Madhu> sure GiovanniMeo . thanks. 17:16:11 <odp-gerritbot> Hsin-Yi Shen proposed a change to ovsdb: BUG-1056 brocastflow is not proper modified after deleting interface https://git.opendaylight.org/gerrit/7270 17:40:14 <rrsantos> hello! can someone give me some lights finding how to disable the module that enables reactive forwarding in opendaylight? I've been using the ovsdb nb api to set up some flows, but some of them are overridden by reactive flows 17:42:38 <CASP3R> I think it's simple forwarding 17:44:04 <odp-gerritbot> A change was merged to ovsdb: Cleaned up Integration Test setup https://git.opendaylight.org/gerrit/7233 17:44:20 <rrsantos> CASP3R: thank you. I'll try it out. By that the way, is there any way for disabling a bundle automatically in odl? 17:44:44 <CASP3R> Not off the top of my head 17:45:17 <networkstatic> ss simple 17:45:35 <networkstatic> you can just delete the simpleforwarding jar buddy 17:45:51 <networkstatic> from the /plugins directory 17:46:02 <CASP3R> yea i know we do that when doing performance testing but didn't know if there a "better' way to do it 17:46:39 <networkstatic> ah roger, proper! 17:46:52 <networkstatic> it shouldnt be bundled in the next integration releases 17:47:05 <networkstatic> enough complained when it was in hydriogen 17:47:24 <rrsantos> networkstatic: so right now I could just just remove it from there and put it back in case I need it 17:50:39 <odp-gerritbot> A change was merged to ovsdb: Fixed a bug in where condition usage in Update & Delete operations. Also moved Delete operation to a seperate transaction in the Test code to test multiple transactions. https://git.opendaylight.org/gerrit/7298 17:50:40 <odp-gerritbot> A change was merged to ovsdb: Commit operation as per Section 5.2.7 of RFC 7047 https://git.opendaylight.org/gerrit/7297 17:52:55 <odp-gerritbot> A change was merged to ovsdb: Abort operation as per Section 5.2.8 of RFC 7047 https://git.opendaylight.org/gerrit/7300 17:53:28 <networkstatic> yessir rrsantos 17:53:35 <networkstatic> ss gives you all of the running modeuls 17:53:37 <networkstatic> modules 17:53:46 <networkstatic> from the OSGI cli 17:54:22 <networkstatic> then u can stop and start based on the PID 17:54:33 <networkstatic> number of the module process in the output 17:54:38 <networkstatic> stop or start etc 17:54:44 <networkstatic> with the number 17:55:34 <rrsantos> networkstatic: yes, it is working now. by the way, should I need to set up a rule for sending arp requests to the controller after I disable the simple forwarding? 17:55:55 <networkstatic> ah ARP lol, thats a toughy 17:56:26 <networkstatic> I need to look at what if anything has been done with arp tracker 17:56:26 <networkstatic> hosttracker 17:56:36 <CASP3R> doesn;t the ARPhandler do that? 17:56:38 <networkstatic> darn, evan isnt on 17:56:50 <networkstatic> roger, simple forwarding loads that 17:57:16 <networkstatic> i know evan had been working on L2switch 17:57:22 <networkstatic> i am just not sure 17:57:33 <networkstatic> we use openstak/ovsdb for ARP 17:57:41 <networkstatic> or to not need arp :) 17:57:58 <networkstatic> the beauty of openstack for networking 17:58:26 <networkstatic> much harder problem discovering w/o orchestration 17:59:14 <networkstatic> rrsantos: soon as colin dixon or evan are on we can sync u up 18:00:27 <rrsantos> networkstatic: thank you :) so a possible solution for handling arp requests after disabling simpleforwarding would be using openstack? 18:01:46 <networkstatic> if u want to give openstack a run here are instructions buddy http://networkstatic.net/updated-devstack-opendaylight-vm-image-for-openstack-icehouse/ 18:01:55 <networkstatic> and a nwe image 18:04:34 <rrsantos> I should give it a try 18:05:03 <rrsantos> I've been reading so much things about openstack but never tried it though 18:10:30 <Madhu> odp-gerritbot: wakeup... 18:10:37 <dave_tucker> ha 18:10:42 <networkstatic> dont u snap at odl_meetbot 18:10:48 <networkstatic> he will cut you 18:11:08 <networkstatic> fiesty little bugger 18:11:45 <Madhu> he is partying too much. 18:11:50 <networkstatic> rofl 18:12:44 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Update .gitreview to use topic/schema branch https://git.opendaylight.org/gerrit/7303 18:13:32 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Stop isReady method always returning true https://git.opendaylight.org/gerrit/7121 18:25:36 <Mierdin> Madhu: how do you suggest I merge the two commits? 18:27:12 <dave_tucker> Mierdin: copy/paste and abandon the other? 18:27:30 <dave_tucker> you could dance with cherry-pick and rebase but it might get confusing quickly 18:30:46 <Mierdin> yeah I've done rebase before and know it's not my first choice :) 18:31:18 <networkstatic> evanz_: hey bud, you guys done anything significant with the L2Switch yet? If so can u dump usage for rrsantos ? 18:34:05 <evanz_> networkstatic: not as far as I am aware 18:34:32 <networkstatic> rogr 18:34:35 <evanz_> the cisco guys are doing cisco live this week, should get more active after that 18:34:57 <networkstatic> rrsantos: looks like simpleforwarding is the out of the box for now buddy 18:35:17 <evanz_> aye 18:35:19 <networkstatic> it basically proxies ARP and does an L3 rewrite 18:35:39 <networkstatic> kinda expensive but does scale beyond L2 segments 18:37:47 <Mierdin> dave_tucker: so the changes are already committed locally 18:37:54 <Mierdin> this is gonna sound stupid 18:37:55 <Mierdin> but... 18:38:14 <bigmstone> Mierdin: there are no stupid questions...unless its from you. 18:38:21 <networkstatic> i can say something dumber if it makes u feel better 18:38:24 <Mierdin> I can maybe undo the Operations changes, commit, then recreate them with a new commit that appends to the commit containing the new Mutate file? 18:38:32 <Mierdin> bigmstone: ass 18:38:38 <dave_tucker> If it's in the same branch.... 18:38:48 <dave_tucker> you can git reset to the change id of your first commit 18:38:54 <bigmstone> networkstatic: that's because we can put a southern draw on it. 18:38:58 <dave_tucker> then git add, git commit --amend 18:39:02 <networkstatic> oh yeah, bigtime 18:39:19 <bigmstone> For reference, I'm sitting next to Mierdin right now. So I'm giving him a hard time. 18:39:20 <networkstatic> maybe some gravy on it too 18:39:32 <bigmstone> Had biscuits and gravy this morning. 18:39:43 <bigmstone> From SF...so it wasn't near as good as it could be. 18:42:08 <rrsantos> networkstatic: I see, thank you. Well in that case is there a way of overriding simpleforwarding flows? Because right now I'm setting up flows with my needs but when some traffic appears a new flow is created by simpleforwarding with the same priority as the previous one 18:42:50 <networkstatic> roger, stopping simpleforwarding and then handcrafting the flows would be the route to go 18:43:25 <networkstatic> look at simpleforwarfing and see how the arp is handled 18:43:44 <networkstatic> u can craft a reply or stitch the broadcast domain out 18:44:03 <networkstatic> or use overlays and not worry about end to end hops in our case 18:44:24 <networkstatic> doesnt apply to physical network yet unfortunately 18:44:39 <networkstatic> but could if u nail up tunnels on ur kit 18:45:14 <networkstatic> u can plumb the ARP request/response flows manually ofcourse 18:45:24 <networkstatic> allow all arp etc 18:45:33 <networkstatic> like a big toughy ACL 18:48:35 <rrsantos> well right now I'm not developing a controller application, so I'm not messing with the ODL code yet. I'm just using the REST API for creating new flows. so the best solution would be to install additional flows allowing all the arp traffic? 18:53:10 <networkstatic> rpger rrsantos that would make sense, also pinging the controller listserv for feedback would be good 18:53:28 <networkstatic> asking for plans on arp handling 18:53:50 <networkstatic> always good to hear needs 18:54:40 <rrsantos> I'm sorry, but what do you mean with pinging the controller listserv? 18:58:18 <networkstatic> sorry, sending an email to controller-users https://lists.opendaylight.org/pipermail/ 18:58:45 <networkstatic> for any dynamic way to handle arp other then simpleforwarding 18:58:51 <networkstatic> or plans in the future 18:59:08 <networkstatic> or hit the main channel, i jsut havent seen anything new on that front 19:06:05 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Unit Test Coverage for OvsdbClientImpl https://git.opendaylight.org/gerrit/7122 19:08:31 <rrsantos> ahhhh got it networkstatic, I'll send a mail to it soon. thank you for all the help! 19:08:46 <networkstatic> great buddy 19:08:48 <networkstatic> u got it 19:13:18 <Mierdin> dave_tucker: here's my mod to OvsDBClientTestIT 19:13:19 <Mierdin> http://pastebin.com/BqbMHfWe 19:13:35 <Mierdin> following Madhu's advice, no need for the set method 19:13:46 <Mierdin> and had to add the Where reference to Mutate.java 19:13:47 <dave_tucker> Mierdin: lgtm 19:14:12 <Mierdin> cool. I'm going to test using your ovs vagrant stuff first and then commit 19:14:20 <Mierdin> which….by the way….is bad ass 19:26:20 <Mierdin> dave_tucker: hmmm doesn't look like docker is installed on the centos VM pulled down by vagrant 19:26:34 <dave_tucker> for which repo sir? 19:26:51 <dave_tucker> docker-ovs? 19:28:07 <Mierdin> yup 19:28:35 <Mierdin> sudo docker pull dave-tucker/docker-ovs:2.1.2 19:28:40 <Mierdin> docker: command not found 19:29:12 <Mierdin> weird cuz I see on your screenshot immediately after vagrant up/reload/ssh 19:29:51 <Mierdin> oops, there was an error with setting up epel repo that scrolled by and I missed it 19:29:56 <Mierdin> looks like a 404 19:29:58 <Mierdin> let me try to rebuild 19:33:54 <Mierdin> dave_tucker: while I'm waiting for that, I'm missing where to specify IP address in OvsDBClientTestIT 19:34:07 <Mierdin> how do you point that test case to the OVS instance set up in docker? 19:34:30 <dave_tucker> yessir. your docker vm will be 192.168.50.4 iirc 19:35:17 <dave_tucker> Madhu_offline's changed merged today on integration test properties 19:35:21 <dave_tucker> let me find the gerrit 19:36:15 <dave_tucker> https://git.opendaylight.org/gerrit/#/c/7233/ 19:37:15 <dave_tucker> Mierdin: so mvn -Dovsdbserver.ipaddress=192.168.50.4 -Dovsdbserver.port=6640 -Pintegrationtest verify 19:38:23 <Mierdin> oh right gotcha I remember that now 19:38:26 <Mierdin> will give that a try 19:38:37 <Mierdin> good work with this vagrant stuff man, it's pretty hands-free 19:38:50 <bigmstone> Mierdin: look at GV. 19:41:42 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Fix TestClient test Case https://git.opendaylight.org/gerrit/7320 20:01:05 <odp-gerritbot> Matt Oswalt proposed a change to ovsdb: Create Mutate Operation https://git.opendaylight.org/gerrit/7321 20:01:18 * Mierdin hugs odp-gerritbot 20:01:56 <Mierdin> I have to run to a session real quick. Hoping to test the addition today but wanted to push this so it wasn't in your way anymore 20:07:42 <Mierdin> bah - still broke. Abandoned for now but the stuff I needed to do should be mostly there 20:08:00 <Mierdin> just too busy right now and can't concentrate. I'll get back on later today and make sure this gets done right 20:08:22 <Mierdin> I'll be on MierdinMobile for rest of day 21:29:01 <Madhu> networkstatic: hey buddy 21:29:12 <networkstatic> yoos 21:29:43 <Madhu> Mierdin: hi. sorry i just came back. will read through the messages. I will also send a proposal on your way. 21:29:52 <Madhu> proposal to fix the Mutate issue :) 21:40:22 <dave_tucker> Madhu:that was a long lunch :p I managed to rebase my commits sir so you are all good to review 21:40:38 <dave_tucker> will be watching odp-gerritbot for anything incoming from you ;) 21:43:51 <dave_tucker> Madhu: and can you also update the trello when you get a sec sir? sure you'd like to check Abort + Commit off! 21:43:57 <Madhu> dave_tucker: lol. ofcourse sir 21:44:14 <dave_tucker> Madhu :p 21:44:21 <Madhu> dave_tucker: did u get a chance to review the RFC7047 ? 21:44:36 <dave_tucker> yessir. we have all the rpc and operations covered 21:44:48 <dave_tucker> in the trello at least 21:45:03 <Madhu> sure. I think we have to address the notifications next 21:45:08 <dave_tucker> checking for any errata or updates 7047 as we speek 21:45:11 <Madhu> today we handle hello, update 21:45:34 <Madhu> we have few notifications not addressed 21:46:19 <Madhu> dave_tucker: once we have the notifications covered, we are in good shape with library. 21:46:32 <Madhu> reasonable shape :) 21:46:52 <dave_tucker> Madhu: I don't see any mention of hello in the rfc 21:47:18 <Madhu> sorry. Echo :) 21:47:27 <dave_tucker> lol. ok with you now :) 21:48:15 <dave_tucker> yeah those are going to be fun :) 21:48:59 <dave_tucker> fixed our issue with the IT sir (port conflict), that's in review... also renamed the class IT as it's most definately an IT not a UT. 21:49:11 * dave_tucker hope's thats ok 21:52:30 <Madhu> dave_tucker: lemme take a look. 21:53:02 <dave_tucker> only if it doesn't interfere with your other commits boss ;) 21:53:11 <dave_tucker> you need a break from code review sometimes 21:56:34 <Madhu> dave_tucker: i know :) at this point, am not even reviewing any code. 21:56:48 <Madhu> am looking at Matt's fix and help massage it a bit. 21:56:58 <dave_tucker> :D 21:57:01 <Madhu> then an important email reply .. followed by code-review ;) 21:58:44 <dave_tucker> you've got it all planned out bro 21:59:03 <dave_tucker> i'm working on converting ABNF -> ANTLR -> Java Classes for Yang support in InteliJ 21:59:07 <dave_tucker> fun times 22:08:37 <Madhu> shague_: networkstatic just replied to an email regarding LBaaS proposal from Chris, uw.edu 22:08:56 <ekarlso> Madhu: what lb ? 22:09:04 <Madhu> since shague_ is looking into it as well, i extended the invite to Chris to participate in the work 22:09:17 <Madhu> ekarlso: u always jump out of the bush and ask questions ;) 22:09:25 <Madhu> https://docs.google.com/document/d/169sTVsXHT7Gdi3F8XRktx1093p5ai2Tl_mQRUfLy24c/edit# 22:09:25 <ekarlso> Madhu: aint I allowed ? 22:09:27 <ekarlso> :p 22:09:28 <Madhu> that lb :) 22:09:29 <networkstatic> excellent, oh yeah, been friends with him for a while 22:09:36 <Madhu> ekarlso: of course no sir :) u r always welcome 22:09:42 <networkstatic> smart dude 22:09:47 <networkstatic> chsmall in here too 22:09:54 <Madhu> ekarlso : u first message starts with a ? always ;) 22:10:09 <Madhu> chsmall: ah. he is here. sorry didn't realize it 22:10:10 <networkstatic> LOL 22:10:12 <networkstatic> hahaha 22:10:17 <networkstatic> first question 22:10:18 <networkstatic> lol 22:10:24 <chsmall> hi 22:10:35 <networkstatic> heya bro 22:10:40 <Madhu> chsmall: didn't realize u r part of this already :) 22:10:49 <Madhu> wouldn't have been formal in replying to the email ;) 22:11:42 <Madhu> chsmall: so glad you sent out that proposal. we were talking about it for a few weeks now and shague_ has picked up to work on it. 22:12:28 <Madhu> ekarlso: if u have any questions... chsmall is here to answer ;) 22:12:35 <networkstatic> hehe 22:12:41 <networkstatic> well played Madhu 22:13:15 * Madhu still not able to shake myself off that stupid email i just sent out without realizing chris is already here in this channel 22:13:26 <chsmall> I chatted with shague_ a bit last week but wanted to put something out there to start the discussion 22:13:41 <Madhu> chsmall: excellent. 22:13:45 <Madhu> sorry didn't realize all these 22:15:09 <networkstatic> Madhu im sorry bro :( my fault, i shoulda made the connetction sooner 22:15:15 <networkstatic> but damn u r so polite 22:15:25 <Madhu> wtf. why polite ? :) 22:15:27 <networkstatic> it will be a reference for future people coming to the project 22:15:31 <Madhu> am glad that someone is doing it 22:15:45 <networkstatic> u r friendly 22:15:54 <networkstatic> we are friendly project 22:16:24 <networkstatic> except for ekarlso 22:16:35 <Madhu> lol. man. 22:16:40 <networkstatic> hehe 22:16:53 <networkstatic> he rates every bit of harrasment he gets :) 22:17:22 <Madhu> he is a lone warrior now sir in this channel 22:17:28 <networkstatic> chsmall i havent had a chance to dig in, soon as i get a stupid service registered and put up for review im digging in! 22:17:33 <networkstatic> haha 22:17:39 <Madhu> we have converted all other ex-colleaques of his to sanity ;) 22:17:55 <dave_tucker> lol 22:18:06 <Madhu> dave_tucker: u got the drift ? ;) lol 22:18:26 <dave_tucker> no the dave_tucker at HP is a right jackass. he rights buggy code 22:18:32 <dave_tucker> have no idea who he is 22:18:45 <chsmall> Just catching up on the e-mail. I think it makes sense to LBaaS in ovsdb since 1) it will be mostly done inside ovs 2) Not sure if I want to spinup a whole project to start with 22:19:41 <Madhu> chsmall: good call. 22:20:09 <networkstatic> and openstack services are the only real thing in SDN today 22:20:11 <networkstatic> oops 22:20:24 <networkstatic> i tried to find a good lone wolf picture for endre 22:20:28 <networkstatic> best i came up with is 22:20:29 <networkstatic> https://www.dropbox.com/s/51dmgdfpytegk8h/Screenshot%202014-05-21%2018.19.30.png 22:20:32 <ekarlso> what u want with me :p 22:20:41 <Madhu> ROFL 22:20:49 <networkstatic> lol 22:20:59 <chsmall> However how LBaaS is done may effect other OpenStack advanced network services so I am hoping to get as much input as I can. 22:21:41 <networkstatic> does the Neutron API differentiate between LBaaS and Port load balancing? 22:21:51 <networkstatic> well, nm, 22:22:00 <Madhu> chsmall: agreed. will review the proposal 22:22:03 <Madhu> and give feedback. 22:22:21 <networkstatic> only thing we care on overall is fulfilling V2 api requests 22:22:22 <Madhu> and also, i heard a lot of interest on LBaaS from EBay folks who attended the OVS hackathon 22:22:30 <Madhu> lets get some inputs form ashw7n as well. 22:22:32 <networkstatic> rest is whatever implementation we want bro 22:22:36 <networkstatic> good call 22:23:00 <networkstatic> chris have u looked at NXMs buddy? 22:23:16 <networkstatic> tcp_flags may be handy for you also 22:23:43 <networkstatic> to avoid a punt fest 22:26:10 <chsmall> There is a proposal about LBaaS from someone at Blue Box. He was working with Ebay/Rackspace who are very interested looking at how to enhance OpenStack LBaaS 22:26:24 <networkstatic> nice 22:26:36 <networkstatic> will check the blueprint 22:27:00 <networkstatic> andyhky is a Racker if u need details 22:27:06 <networkstatic> and a hell of a guy 22:27:14 <networkstatic> who randomly lives 20 minuutes from me 22:29:27 <chsmall> Only mentioned extensions in passing in the strawman. Very useful but they are extensions and you have to be careful about relying too much on them. 22:32:23 <chsmall> Would like to have some level of graceful fallback to lowest denominator to standards but add more functionality/performance if you have the correct resources 22:33:19 <networkstatic> rogr boss, for sure 22:33:24 <networkstatic> we used to be that way 22:33:46 <networkstatic> until we realized we couldn be as useful 22:33:50 <networkstatic> now we are arms dealers 22:34:15 <networkstatic> gimme onePk, JunOSSDK, EOS, whatever it takes 22:34:20 <networkstatic> hehe, j/k 22:34:38 <networkstatic> ONF being reluctant to have a SW spec is frustrating 22:35:03 <networkstatic> maybe u can push :) ur a researcher n ONF arent u? 22:40:11 <chsmall> You definetly need to use extensions or reach-arounds (as IU has coined the term) to do useful things. I having been pushing for more useful info in 22:40:53 <chsmall> capabilities since somewhere around OF 0.8.9 22:42:21 <Madhu> chsmall: agreed. 22:42:23 <chsmall> If you knew exactly what the switch was capable of you could build a strategy to implement the task. 22:43:03 <Madhu> when it comes to OVS at least :) depending on NXM seems reasonable 22:43:36 <Madhu> and as networkstatic said, we were originally an OXM only shop to realize that we cannot do everything that we want to 22:43:56 <Madhu> also from the history, we can see that NXMs get into OXM and OF spec but a bit delayed 22:44:21 <networkstatic> so tru, TLVs in OVS are implemented the same 22:44:27 <networkstatic> only diff is the constant sent 22:44:44 <Madhu> also, if we have to over-simply or grossly-incorrect :) we can see OVSDB as an extension as well :) 22:45:03 <networkstatic> and Im a hack 22:45:08 <Madhu> it extends the Flow functionality to configuration and both work hand in hand very well 22:45:11 <networkstatic> not even an extension 22:45:53 <Madhu> of course all these are true for OVS :) 00:04:05 <networkstatic> hell 00:04:27 <networkstatic> random 00:25:25 <evanz> networkstatic: is it madhu's birthday? I got a g+ notif 00:26:07 <tbachman> Madhu_offline: Happy b-day man! 00:26:18 <evanz> woo! 00:42:38 <MierdinMobile> sorry for the condition I left my stuff in. Crazy out here. Just trying to keep stuff moving 00:43:29 <MierdinMobile> The vagrant stuff is wicked cool though I will get back to it ASAP 03:10:24 <Madhu> evanz: tbachman thanks guys. yes 03:10:46 <Madhu> its like a biological clock for me. tic toc tic toc ;) 03:12:07 <evanz> :) 04:01:21 <Madhu> ashw7n: 04:01:28 <ashw7n> yes boss 04:01:53 <Madhu> few questions sir 04:02:19 <ashw7n> sure.. 04:03:27 <Madhu> we are working on the schema branch as u might have noticed 04:03:58 <Madhu> a lot of questions actually :) but i will restrict to a few 04:04:23 <Madhu> in the master we used the OvsDBMap and OvsDBSet 04:04:39 <ashw7n> yes 04:04:55 <Madhu> which does the nasty ovsdb json stuff such as including "set" and "map" for the multi-dimensional array 04:05:41 <Madhu> in the current monitor code 04:06:16 <Madhu> am looking at the response handling portion especially 04:06:22 <Madhu> MonitorHandle monitor = ovs.monitor(dbSchema, monitorRequests, new MonitorCallBack() { 04:06:31 <Madhu> in OvsDBClientTest 04:06:57 <Madhu> the MonitorCallback seems to use the TableUpdates from the past 04:06:58 <ashw7n> checking... 04:07:10 <ashw7n> yes, we need to get to the bottom :) 04:07:26 <ashw7n> what we need is a simple approach just like what we have for the schema 04:08:48 <Madhu> am still confused sir :) 04:09:13 <Madhu> how are the ovsdb json "set" and "map" translated now to a Java Set and a Java Map ? 04:09:27 <Madhu> since we are not explicitely using ovsdbmap and ovsdbset 04:10:04 <ashw7n> public static DatabaseSchema fromJson(String dbName, JsonNode json) { 04:10:10 <ashw7n> in DatabaseSchema 04:10:56 <ashw7n> wait, i didnt get the part of ovsdb set thing, we havent done that yet 04:11:31 <ashw7n> we need to change the mapping either manually, just like the DatabaseSchema, or use a generic translation like we have on the existing code 04:12:09 <Madhu> if u look at our master 04:12:28 <ashw7n> okay 04:12:36 <Madhu> tables like Bridge, etc.. are all defined with OvsSet and OvsMap 04:13:46 <Madhu> this helps in ser and deser back and forth between the java world and the "set", "map" ovsdb json world 04:14:33 <Madhu> am just trying to see how we will be addressing it in the new schema driven world 04:14:50 <Madhu> i hope we will expose it as a simple Set or Map in Java 04:15:00 <Madhu> without the need to expose OvsSet and OvsMap 04:15:01 <ashw7n> yes, it needs to be a simple Set and Map in java 04:15:25 <ashw7n> and our serialization logic would have to deal with it, based on the Schema definition 04:15:38 <Madhu> got it. 04:15:40 <ashw7n> or the spec 04:15:54 <ashw7n> but the good part is that we have a schema which has all the details and types 04:15:59 <ashw7n> so we shoudl be able to do it 04:16:01 <Madhu> yes. 04:16:08 <Madhu> actually we have to cleanup the schema branch 04:16:13 <Madhu> and remove all the old references sir 04:16:18 <ashw7n> i will work on one example tomorrow 04:16:20 <Madhu> it is getting damn confusing seeing all here 04:16:22 <ashw7n> yes we should 04:16:32 <ashw7n> maybe we can spent some time tomorrow together 04:16:40 <Madhu> yes sir. 04:16:55 <Madhu> few of us are filling in the missing operations 04:17:10 <Madhu> adding tests gets challenging because we don't have these coveree 04:17:29 <Madhu> because without the ovsdbset and ovsdbmap covered 04:17:36 <Madhu> we cannot even handle monitor 04:17:41 <Madhu> without monitor we cannot do mutate 04:17:49 <Madhu> and the chain continues :( 04:19:03 <ashw7n> ohh, okay, so can you tell me one operation which is using set and map and i'll work on it 04:19:22 <Madhu> oh. 04:19:27 <Madhu> monitor as an example sir 04:19:49 <Madhu> once we set Monitor. the monitor response (and the update) will return data with this stupid json format 04:20:01 <Madhu> that is the most important of it all. 04:20:23 <Madhu> the other is the insert, update, mutate operations. 04:20:29 <Madhu> since insert is already implemented. 04:20:38 <Madhu> we have to make Insert work for all the set/map cases as well 04:21:09 <ashw7n> okay, so what is the TableUpdate returning now, sorry its been couple weeks :D 04:21:30 <Madhu> i will give u an example sir 04:21:58 <Madhu> https://gist.github.com/5c194ad56fa4a038a69e 04:22:25 <Madhu> this is an example where a monitor is done on Bridge table for columns name, fail_mode and protocols 04:22:31 <Madhu> name, fail_mode are both simple strings. 04:22:38 <Madhu> but protocols is a set 04:22:43 <ashw7n> ok 04:23:24 <Madhu> the return as u can see has the ----- protocols":["set",["OpenFlow10","OpenFlow13"]] 04:23:40 <Madhu> this was originally mapped to OvsDBSet 04:23:51 <Madhu> and mapped statically to the protocols object in the Bridge Class 04:24:01 <ashw7n> lol, I know this, but what i was asking is about the current method we have for getUpdate 04:24:15 <ashw7n> there is a test case in OvsDbClietnTestIT 04:24:20 <ashw7n> for monitor 04:24:25 <ashw7n> isnt it returning 04:24:25 <ashw7n> ? 04:24:41 <ashw7n> I thought it was working before I checked in 04:24:50 <ashw7n> testMonitorRequest 04:24:55 <Madhu> ya i saw that :) but the TableUpdates didn't have toString() :) 04:25:03 <Madhu> and hence the output is this : result = org.opendaylight.ovsdb.lib.message.TableUpdates@2f2b69ef 04:25:05 <Madhu> lol 04:25:17 <Madhu> am adding the toString now to check what the result is parsed out as 04:25:30 <Madhu> but as u can see, the result will be parsed out properly because 04:25:56 <Madhu> TableUpdates is statically typed class for the static Tables (from the existing code) 04:26:03 <Madhu> so we have to stop using this :) 04:26:28 <ashw7n> okay, i think it will be much better if we can spent some time tomorrow 04:26:45 <ashw7n> and we can hack it out 04:28:26 <Madhu> ok sir 04:30:00 <Madhu> ashw7n: shall we do hangout tomorrow to begin with and then I can even drive in to your office 04:30:58 <ashw7n> okay 04:31:29 <Madhu> ok sir. thanks. 04:31:47 <Madhu> the old code and new code in the same folder causes a lot of confusion 04:32:00 <Madhu> am planning to create a new package called old and move the old classes into it 04:32:18 <Madhu> (to avoid compilation issues on the application layer (plugin and neutron) 04:36:04 <ashw7n> i left the old ones in there so taht we dont break everything 04:36:10 <ashw7n> i told you this was a big change :D 04:36:37 <Madhu> of course sir. 04:37:02 <Madhu> we are on path for helium release now :) and this is getting HOT 04:46:09 <Madhu> ashw7n: just trying my luck :) do u have sometime tonight ? 04:46:23 <Madhu> for a hangout in 30 mins ? 04:46:34 <ashw7n> sure, in 10 minutes? 04:46:56 <Madhu> 15 :) my kid shud be asleep by then. 05:00:03 <ashw7n> okay 05:00:05 <ashw7n> ping me please 05:05:20 <Madhu> ashw7n: am ready sir. will ping u in hangout 12:02:09 <Madhu> dave_tucker: minor nit on select fix 12:03:01 <dave_tucker> Madhu: hey buddy 12:03:24 <dave_tucker> Madhu: this one is rebased now sir https://git.opendaylight.org/gerrit/#/c/6861/ 12:04:04 <Madhu> yes. 12:04:16 <Madhu> the method name "value" in the Select class may not be appropriate 12:04:42 <dave_tucker> ok. I'm playing follow the leader here - as I said, copy/paste ninja 12:07:14 <Madhu> dave_tucker: trying to find a more appropriate method name from the existing example 12:07:53 <Madhu> dave_tucker: the closest i could find is the addColumn method of Monitor 12:09:56 <dave_tucker> Madhu: would "row" be better than "value"? 12:10:49 <dave_tucker> or rows rather... or even just select ;) 12:11:08 <dave_tucker> From the RFC... The operation searches "table" for rows that match all the conditions 12:11:08 <dave_tucker> specified in "where". 12:11:23 <Madhu> yes. 12:11:30 <Madhu> but rows is not the right one here 12:11:33 <Madhu> it is the "result" object 12:11:45 <Madhu> i think addColumn is the most appropriate 12:11:54 <Madhu> and it should be similar to monitorrequest.java 12:11:57 <dave_tucker> but wouldn't it be addRows 12:12:03 <dave_tucker> or addRow 12:12:09 <Madhu> nope. 12:12:11 <dave_tucker> rather.... as the result of a select is Rows 12:12:16 <Madhu> Select is not adding a row 12:12:19 * dave_tucker is missing something 12:12:46 <Madhu> it is searching for data in a Table given a where condition and columns of data to Select from 12:12:52 <Madhu> and the result object is the row. 12:12:59 <Madhu> rather rows 12:13:18 <Madhu> so the "value" method is not appropriate. and must be replaced with addColumn 12:13:36 <Madhu> and the test case must include that. (now it is ignored and uses just the where clause) 12:15:09 <dave_tucker> we have a setColumns method that adds columns to the list... how does addColumns differ? 12:15:22 <dave_tucker> not trying to be awkward... just struggling to understand 12:17:03 <Madhu> of course sir :) 12:17:12 <Madhu> setColumns is the typical getter setter. 12:17:37 <Madhu> addColumn() is a usage aspect of the API 12:18:28 <dave_tucker> ok.. so other than the name is the method signature ok? 12:18:29 <Madhu> we are providing a choice to the user to choose to write addColumn() one after another instead of a precooked list. 12:18:34 <dave_tucker> generics confuse the hell out of me 12:18:36 <dave_tucker> :S 12:18:57 <Madhu> actually the method signature is also not appropriate :( 12:19:09 <Madhu> because. there is no "value" per-se 12:19:22 <Madhu> and adding it to the rows is also not correct 12:19:27 <Madhu> it should be adding to the columns 12:19:42 <Madhu> because rows is the return value of select operation 12:19:48 <Madhu> not something set by the user 12:19:49 <dave_tucker> public <D, C extends TableSchema<C>> Select<E> addColumns(ColumnSchema<C, D> columnSchema) 12:20:03 <dave_tucker> then columns.put(columnSchema) 12:20:19 <dave_tucker> and remove private Map rows as it's the return value 12:20:21 <dave_tucker> ? 12:21:10 <Madhu> please refer to MonitorRequest.java 12:21:16 <Madhu> it does exactly the same that we are looking for 12:21:23 <dave_tucker> https://gist.github.com/a89bcd35f250158690e6 12:21:35 <dave_tucker> I have that class open too Madhu :) 12:23:01 * dave_tucker changes gistify to gist in java 12:24:23 <dave_tucker> Madhu: highlighting in java https://gist.github.com/anonymous/b78271c0967dc3d192ad 12:25:01 <Madhu> s/addColumns/addColumn/ 12:25:07 <dave_tucker> oops! 12:26:12 <dave_tucker> done. grabbing coffee brb 12:32:39 <dave_tucker> ok back in the room 12:34:37 <dave_tucker> Madhu: would it make sense to just call the method "column" 12:35:11 <dave_tucker> .add(op.select(bridge).column(name) 12:35:14 <Madhu> dave_tucker: was actually thinking about that 12:35:15 <Madhu> but 12:35:23 <Madhu> method() is an action 12:35:28 <Madhu> and column is a noun :) 12:35:31 <Madhu> not a verb 12:35:41 <Madhu> thats y addColumn makes more sense 12:36:01 <dave_tucker> well where is an adverb 12:36:06 <dave_tucker> :p 12:36:07 <Madhu> lol. 12:36:37 <dave_tucker> but from a usage standpoint, you are saying "select" "column" not "select" "addColumn" 12:36:39 <Madhu> for other places such as set(), mutate(), abort() 12:37:11 <Madhu> these are all actions and hence update has a set(), etc... 12:37:25 <Madhu> ya. select is the verb there :) 12:37:39 <Madhu> and column is the data / noun that goes in as parameter 12:38:02 <dave_tucker> what about "from" 12:38:04 <dave_tucker> like sql 12:38:11 <dave_tucker> select * form 12:38:15 <dave_tucker> s/form/from 12:38:20 <Madhu> actually no :) 12:38:27 <Madhu> we are selecting the columns 12:38:39 <Madhu> from is for the table to select from 12:38:46 <Madhu> the * is what is being discussed here 12:39:25 <dave_tucker> crap! 12:39:29 <Madhu> as per the RFC. the Select Column is exactly the same as Monitor Column 12:39:36 <Madhu> thats y i chose that example 12:39:58 <Madhu> Select : "columns": [<column>*] optional 12:40:15 <Madhu> Monitor-request : "columns": [<column>*] optional 12:40:20 <dave_tucker> I know... I know... but I like the notation we have 12:40:25 <dave_tucker> .column is pretty 12:40:32 <dave_tucker> .addColumn is not 12:40:36 <dave_tucker> :p 12:40:49 <Madhu> if u can verbify column() then am with u there ;) 12:41:16 <dave_tucker> I'm going to column get you Madhu 12:41:18 <dave_tucker> there done 12:41:21 <dave_tucker> lol 12:41:32 <Madhu> lol. nice try 12:42:38 <Madhu> but if u feel column() is more usable and intuitive 12:42:46 <Madhu> its okay. am not trying to be a poet here ;) 12:42:49 <dave_tucker> rofl 12:43:03 <dave_tucker> lets put it to the vote... otherwise its just my say vs yours 12:43:16 <dave_tucker> we need a mediator lol 12:43:33 <dave_tucker> where's odl_meetbot 12:44:31 <dave_tucker> .add(op.select(bridge).column(colA).column(colB).where(colA.x = "y") 12:44:33 <dave_tucker> vs 12:44:41 <dave_tucker> .add(op.select(bridge).addColumn(colA).addColumn(colB).where(colA.x = "y") 12:44:59 <Madhu> of course #1 is more intuitive :) 12:45:22 <dave_tucker> right I'm rolling with #1 before you change your mind lol! 12:45:24 <Madhu> because it sounds just like english here 12:45:48 <dave_tucker> yeah, tbh I think we can excuse it not being a verb because it's being used to construct an operation 12:45:50 <Madhu> select from bridge column colA, colB where colA.x=y 12:46:03 <Madhu> #agreed ;) 12:46:24 <dave_tucker> lol 12:46:28 <Madhu> okay sir. lets do it. 12:46:38 <dave_tucker> same for MonitorSelect 12:46:40 * dave_tucker ducks 12:47:33 <Madhu> well... there i have some problem :) 12:47:44 <Madhu> because it is a "Builder" 12:47:54 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Add OVSDB Select Operation https://git.opendaylight.org/gerrit/6861 12:47:55 <Madhu> not sure why we have 2 different patterns here though 12:47:59 <tbachman> qq — this all makes sense b/c the context here is add 12:48:02 <Madhu> Ashwin can help answer it for us 12:48:10 <tbachman> Will this ever be used w/o the add, or other verb operation? 12:48:18 <tbachman> B/c then the column may not make sense? 12:48:21 * tbachman ducks 12:48:34 <Madhu> tbachman: good question. 12:48:38 <dave_tucker> tbachman: I don't think so 12:48:46 <Madhu> but in this. it is used only for add 12:48:48 <dave_tucker> there are normal getter/setters for use outside 12:48:50 <dave_tucker> I think 12:49:03 <tbachman> so, that seems reasonable :) 12:49:06 <Madhu> because RFC doesn't indicate anything like exclude :) 12:49:39 <Madhu> if there is an exclude option in select column 12:49:50 <Madhu> then i agree.. we are in trouble with the simplicity of column() :) 12:50:52 <dave_tucker> but it looks so cool! it's like it's own DSL 12:52:48 <Madhu> they say programming is an art :) 12:52:57 <Madhu> and with today's poetic discussion, we proved it. lol. 12:53:51 <tbachman> Madhu: question for you on OVSDB 12:54:03 <Madhu> oops. am in trouble now 12:54:10 <tbachman> LOL 12:54:22 <tbachman> no — I *should* actually know the answer, but don’t 12:54:36 <tbachman> You created some extra code for “bi-directional” JSON RPC 12:54:38 * dave_tucker lines up the next question for Madhu 12:54:45 <tbachman> (heard you mention that a few times) 12:54:54 <tbachman> Were there no libs out there to do this? 12:55:02 <tbachman> Or just none that provided what u needed? 12:55:11 <Madhu> #2. 12:55:16 <Madhu> none that provided what we need 12:55:20 * tbachman steps aside to let dave_tucker ask his q 12:55:24 <tbachman> Madhu: got it. 12:55:42 <tbachman> Reason I’m asking — will likely be doing the SB plugin 4 ur fave prototocol ;) 12:55:44 * Madhu thinks tbachman is ready to reuse some of it for his "project" 12:55:52 <tbachman> lol 12:55:55 <tbachman> read my mind 12:56:03 <Madhu> and thats y i hate that. 12:56:09 <tbachman> yeah 12:56:17 <dave_tucker> Madhu: you never answered my question about ServiceHelper. why we use that vs just using OSGi to inject services to the classes that need them :D 12:56:18 <Madhu> because it is going to be a r.i.p.o.f.f ;) 12:56:27 <dave_tucker> Madhu: tbachman: lol 12:56:29 <tbachman> lol 12:56:30 <Madhu> ah. right dave_tucker 12:56:43 <tbachman> Madhu: would you expect anything less from me ;)? 12:56:49 <Madhu> service helper is useful for those bundles who don't have any Activator 12:56:58 <Madhu> tbachman: u as a friend no problem :) 12:56:58 <dave_tucker> tbachman: Ctrl + C, Ctrl + V 12:57:10 <tbachman> dave_tucker: lol! 12:57:14 <tbachman> that’s what I do best :) 12:57:30 <dave_tucker> Madhu: cool, but if we *have* an activator, we should use it right ;) 12:57:40 <Madhu> tbachman: heard another great addition to ur team ;) 12:57:48 <tbachman> ? 12:57:49 * dave_tucker refers to copious usage of ServiceHelper in ovsdb.neutron 12:57:51 <Madhu> dave_tucker: well it depends actually 13:01:01 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Add OVSDB Select Operation https://git.opendaylight.org/gerrit/6861 13:08:15 <dave_tucker> 6861 is ready for merging now Madhu ;) before you change your mind about columns ;) 13:09:27 <Madhu> lol. 13:09:44 <Madhu> but. it is certainly way better than value() and updating rows. do u agree dave_tucker ;) 13:09:58 <dave_tucker> Madhu: yessir 13:10:09 <dave_tucker> that's what code review is for 13:10:31 <Madhu> certainly sir. glad we are doing it 13:10:33 <dave_tucker> and it forced me to understand what I was doing better, so thanks Madhu :) 13:17:34 <odp-gerritbot> A change was merged to ovsdb: Add OVSDB Select Operation https://git.opendaylight.org/gerrit/6861 13:18:13 <Madhu> dave_tucker: now to the isReady() debate ;) 13:18:24 <dave_tucker> Madhu: lol 13:20:12 <dave_tucker> omg lol 13:20:42 <dave_tucker> lets do https://git.opendaylight.org/gerrit/#/c/7320/ first 13:22:11 <Madhu> dave_tucker: okay :) 13:22:20 <Madhu> am torn between UT vs IT here :) 13:22:31 <Madhu> yes. it is IT :) 13:22:46 <Madhu> but it is a clear UT by not mocking the server portions 13:23:05 <Madhu> if we see Netty library as something that just works like a Java system library 13:23:24 <Madhu> then it is a mocked form of server. won't u agree ? 13:25:59 <dave_tucker> Madhu: I'm torn too 13:26:23 <dave_tucker> but UT should focus on testing logic in a single class 13:26:42 <dave_tucker> this seems to test multiple classes (like an IT) and also uses data that looks real iirc 13:27:43 <Madhu> well yes. thats y am trying to duck with the fact that if we assume Netty as something solid 13:27:49 <Madhu> and the data is not real :) 13:27:57 <Madhu> it is of course faked from a file 13:28:25 <Madhu> does this sound real to u dave_tucker : "planet of apes" ? ;) 13:29:01 <dave_tucker> Madhu: yessir. it's in the SciFi schema in RFC1999 13:29:10 <dave_tucker> j/k 13:29:12 <Madhu> lol 13:29:26 <Madhu> as u can see 13:29:45 <Madhu> this UT was written to test JsonRPCDecoder class 13:29:57 <Madhu> of course with the help of Netty 13:30:18 <Madhu> dave_tucker: it doesn't test anything else. 13:30:45 <dave_tucker> Then it should be JsonRPCDecoderTest ;) and we should be testing those methods 13:31:04 <Madhu> that I agree. 13:31:18 <dave_tucker> starting up servers etc... is normally done in an IT 13:31:19 <Madhu> calling this an IT is a stretch 13:31:26 <Madhu> i agree there too 13:31:35 <Madhu> just that it is an intelligent mocking :) 13:31:36 <dave_tucker> the "right" way to do a UT would be to inject that data directly in to the method being tested 13:31:51 <dave_tucker> Madhu: too much intelligence there ;) 13:33:41 * dave_tucker wonders if there is a mocking framework for netty 13:33:51 <dave_tucker> the netty dependency makes this an IT imo 13:33:59 <Madhu> dave_tucker: i will keep this open for Ashwin to comment 13:34:11 <Madhu> and I spoke to him yesterday 13:34:16 <Madhu> and we are going to fix some issues today. 13:34:19 <dave_tucker> COOL! :D 13:34:24 <Madhu> i identified a few for his taste :) 13:34:39 <dave_tucker> ok the are you ready for isReady :D 13:34:52 <Madhu> i gave my comment already. u decide sir 13:35:54 <dave_tucker> Madhu: java q. does the while break as soon as the condition is met? or does it finish the loop an re-evaluate the condifion before next run? 13:36:32 <dave_tucker> and if I add a "break" inside the if, will that break out of the "if" or the "while" or both ;) 13:36:54 <Madhu> breaks the loop :) 13:37:27 <dave_tucker> ok Madhu... will add a break and call it quits. as I said, I like to return only at the end of a method as I find it easier to read 13:37:31 <Madhu> dave_tucker: http://stackoverflow.com/questions/4154553/the-command-break-in-java-what-if ;) 13:37:48 <dave_tucker> Madhu: thanks sir! 13:38:08 <Madhu> dave_tucker: if u break, then u have to use that stupid variable ;) to know what is the return value 13:38:37 <dave_tucker> lol. hey don't go calling "boolean ready" stupid - he's my friend 13:39:05 <Madhu> ah. i had just 1 freedom to call someone stupid.. now even that is gone. 13:39:45 <dave_tucker> Madhu: you have no freedoms sir. you must suffer my code reviews all day long! 13:40:18 <Madhu> lol. ok buddy 13:40:34 <Madhu> u decide on the usage of boolean vs returning from within ;) 13:41:41 <dave_tucker> Madhu: ha I win! I will wear you down with my amateur java lol 13:42:37 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Stop isReady method always returning true https://git.opendaylight.org/gerrit/7121 13:42:55 <dave_tucker> Madhu: I went with your method though 13:43:24 <dave_tucker> have a feeling that my hatred of returning values mid-method is as irrational as my hatred of Hugh Grant and Colin Firth 13:44:36 <Madhu> ah. that is universal then. i thought it was just me ;) 13:44:46 <dave_tucker> :) 13:44:48 <Madhu> i mean the actor hatred part 13:44:52 <dave_tucker> I got that 13:48:23 <odp-gerritbot> A change was merged to ovsdb: Stop isReady method always returning true https://git.opendaylight.org/gerrit/7121 13:50:43 <andyhky> do we need to get a fund together so networkstatic can have an IRC bouncer? 13:51:35 <odp-gerritbot> Giovanni Meo proposed a change to ovsdb: ovsdb has a dependency on artifacts from integration project, that shall be removed https://git.opendaylight.org/gerrit/7334 13:52:01 <odp-gerritbot> Dave Tucker proposed a change to ovsdb: Unit Test Coverage for OvsdbClientImpl https://git.opendaylight.org/gerrit/7122 13:52:06 <Madhu> andyhky: lol. 13:53:46 <Madhu> andyhky: earthfirewindwaterheartbyyourpowerscombinediamcaptainpla.net thats something. 13:54:12 <andyhky> I try :) 13:54:32 <Madhu> pla.net.... u nailed it 13:56:58 <dave_tucker> Madhu: after GiovanniMeo's changes have merged do you think we can attempt a merge of master to topic/schema... so we get distribution + vagrant + other fixes in our topic branch 13:57:15 <Madhu> dave_tucker: good idea 13:57:26 <Madhu> we need to pick up all the changes going into neutron layer as well 13:57:54 <Madhu> dave_tucker: would u like to give GiovanniMeo's fix a try ? 13:58:00 <dave_tucker> d/ling it now ;) 13:58:06 <Madhu> tx bro 14:00:12 <GiovanniMeo> thx guys, i have done basic validation on it 14:00:16 <GiovanniMeo> as basic as making sure 14:00:21 <GiovanniMeo> the bundles all startup 14:00:26 <GiovanniMeo> and the ovsdb is running 14:00:34 <GiovanniMeo> but nothing more than that 14:00:54 <GiovanniMeo> so any real run is very much appreciated 14:01:17 <GiovanniMeo> also any suggestions on how to avoid caching run.base.sh 14:01:19 <GiovanniMeo> is welcome 14:01:53 <dave_tucker> GiovanniMeo: assume this is not your fault, but get this when starting the controller. https://gist.github.com/45f60f6a63a9953159ec 14:02:28 <GiovanniMeo> i didn't see it in my run 14:02:35 <GiovanniMeo> but i suspect 14:02:36 <dave_tucker> ovsdb bundles start fine 14:02:41 <GiovanniMeo> it's a race condition 14:02:58 <GiovanniMeo> simply thee remoterpc 14:03:12 <GiovanniMeo> should be waiting till the mandatory dependency 14:03:18 <GiovanniMeo> on routing table 14:03:24 <GiovanniMeo> is satisfied 14:03:27 <GiovanniMeo> or should run 14:03:32 <GiovanniMeo> assuming that could be null 14:03:43 <GiovanniMeo> the usual problem of dependency management 14:03:51 <GiovanniMeo> we have seen spread around mdsal :( 14:04:08 <dave_tucker> GiovanniMeo: yeah :) will raise a bug :) but will +1 your change as it lgtm 14:04:41 <Madhu> GiovanniMeo: sorry didn't try the change yet 14:04:45 <Madhu> but easy to ask the question 14:04:46 <Madhu> :) 14:05:09 <Madhu> so we still have the option to use -of13 and choose between ad-sal OF plugin and MD-SAL OF plugin ? 14:08:26 <GiovanniMeo> i just copied 14:08:31 <GiovanniMeo> run.base.sh 14:08:35 <GiovanniMeo> as it's 14:08:41 <GiovanniMeo> so the flag is in there 14:08:50 <GiovanniMeo> didn't tested though 14:11:28 <dave_tucker> Madhu: tested. it works 14:12:07 <Madhu> thanks guys. 14:12:27 <dave_tucker> Madhu: proof - https://gist.github.com/anonymous/a3ba57675ce8c4035fe9 ;) 14:12:44 <Madhu> excellent :) 14:13:00 <Madhu> okay, 7am is here. and time to exit for 90 mins 14:13:36 <dave_tucker> bye Madhu 14:14:02 <Madhu> dave_tucker: before I leave, will push a test code for ur review :) 14:14:14 <dave_tucker> Madhu: ofc, I owe you a review or 100 14:18:29 <odp-gerritbot> Madhu Venugopal proposed a change to ovsdb: Added a test-case to include non-atomic column to test the ser/de-ser pipeline in the new schema base code. https://git.opendaylight.org/gerrit/7335 14:18:49 <Madhu> dave_tucker: all yours sir. bfn 14:19:34 <Madhu> dave_tucker: pls look out for the ---- "row":{"flood_vlans":["set",[100,200]] in the message sent to the OVS 14:19:46 <Madhu> the ugliness of "set" usage is being tested here. 14:19:49 <dave_tucker> ok boss will do ;) 14:20:03 <Madhu> pls look for <set> definition in the RFC and u will understand what am talking about :) 14:20:04 <Madhu> bye 14:33:49 <GiovanniMeo> thx dave_tucker for the test 14:33:54 <GiovanniMeo> much appreciated it! 14:34:05 <dave_tucker> GiovanniMeo: any time! 14:36:03 <edwarnicke> Guys... what is the current version of ovsdb for stable/hydrogen ? 14:36:36 <edwarnicke> (I ask because I happen to be updating the integration stable/hydrogen pom files versions for openflowplugin and noticed ovsdb also appears to be out of date) 14:37:06 <jordi_cttc> guys... although having ping between VM's, why am I seeing red ports in opendaylight? 14:37:33 <jordi_cttc> I am trying to understand how the integration between opendaylight and openstack works in fedora 20 14:38:52 <jordi_cttc> for instance the qg interfaces of br-ex of the openstack control node appears in red 14:39:01 <jordi_cttc> in opendaylight 14:43:55 <jordi_cttc> after stacking, googling and praying, my concerns come because I do not know what do I have to see in the opendaylight GUI 14:44:00 <jordi_cttc> arrgghhhh :S 14:44:46 <jordi_cttc> does anybody know what do we have to see in the opendaylight GUI? 15:45:48 <Madhu> jordi_cttc: hi 15:47:19 <Madhu> jordi_cttc: the open daylight GUI displays what is learnt on the backend 15:47:42 <Madhu> when you say node appears in red ... do u mean the ports attached to the switch appears in red ? 15:48:04 <Madhu> that typically means that the port is down 16:39:03 <Madhu> edwarnicke: did u send any email about ovsdb stable/hydrogen version issue in integration branch ? 16:39:11 <Madhu> i remember seeing it. but now I don't :( 16:39:16 <Madhu> can u pls resend it if u don't mind 18:16:00 <andyhky> networkstatic: i know you missed this. 18:16:02 <andyhky> [09:50:43] <andyhky> do we need to get a fund together so networkstatic can have an IRC bouncer? 18:16:11 <networkstatic> haha 18:16:21 <networkstatic> whatup buddy 18:16:31 <networkstatic> u have a good time last week? 18:33:45 <andyhky> of course! 18:33:48 <andyhky> you? 18:39:23 <abhijitkumbhare> Madhu: or networkstatic: question about Trello. Trying to add this for Openflowplugin. I see you have on the top level ovsdb (https://trello.com/ovsdb) and under that you have the boards Helium, Hydrogen, etc. So is this top level ovsdb an organization? 18:39:49 <networkstatic> roger, buddy 18:39:56 <networkstatic> we had it RO 18:40:02 <networkstatic> but i think its RW now 18:40:21 <Madhu> abhijitkumbhare: yes. 18:40:23 <networkstatic> inside Helium is the current cadence we are on 18:40:30 <abhijitkumbhare> OK - I just created first OpenFlow plugin as a board - will delete and create it as an org 18:40:41 <abhijitkumbhare> got it 18:40:45 <Madhu> abhijitkumbhare: yes. 18:40:51 <Madhu> create an org 18:40:55 <Madhu> and have control over the org 18:41:04 <Madhu> and create boards and share it with others 18:41:05 <abhijitkumbhare> sounds good 18:41:07 <networkstatic> awesome, link us when u get done if you dont mind, would be great 18:41:12 <abhijitkumbhare> sure 18:41:24 <networkstatic> thanks buddy, 18:42:56 <networkstatic> i like trello over bugzilla for roadmap projects 18:43:16 <networkstatic> i really hate searching in bugzilla 18:43:46 <Madhu> if u r an agile focused project, we also use agilefant :) 18:43:52 <networkstatic> searching in Evernote is worse 18:44:02 <Madhu> pretty good free agile sprint tracking tool 18:44:13 <networkstatic> i loved what akim did with toolkit 18:44:16 <networkstatic> so sexy 18:44:31 <networkstatic> not akim but akim's agilefant 18:44:38 <networkstatic> not that he isnt pretty 18:44:42 <networkstatic> what? 18:44:42 <Madhu> lol 18:44:44 <networkstatic> lol 19:08:08 <dfarrell07> To jump in a little late, I've been using GitHub's issues for a large robotics project for quite some time. Quite a fan. Very powerful, love being able to link to commits/files/branches/LoC, good syntax highlighting of code chunks, @tagging for devs and teams, milestones for sprints...I could go on and on lol 19:19:02 <networkstatic> heya dfarrell07 i just realized that u r Daniel! 19:19:04 <networkstatic> lol 19:19:13 <networkstatic> robotics tipped me off 19:23:42 <dfarrell07> networkstat: Yeah lol, dfarrell07 -> Daniel Farrell, SDN intern at Red Hat 19:24:58 <dfarrell07> I know networkstat -> Brent from all the helpful blogs, lol 20:14:47 <abhijitkumbhare> Wow - this Trello stuff is real cool 20:15:27 <abhijitkumbhare> I have added Madhu and networkstatic to the members of OpenFlow plugin Helium release 20:15:41 <Madhu> abhijitkumbhare: thanks. 20:15:42 <abhijitkumbhare> Not assigned any tasks :-) 20:15:56 <abhijitkumbhare> I mean to you guys :-) 20:16:00 <Madhu> lol 20:16:11 <Madhu> don't worry we are crazy enough to pick some ;) 20:16:18 <abhijitkumbhare> sure :-) 21:34:23 * Mierdin finally grabs some time to fix my screw-up from yesterday 21:59:01 <Madhu> Mierdin: hi. 21:59:17 <Mierdin> hey Madhu 21:59:19 <Madhu> lemme know if u need any help sir 21:59:26 <Mierdin> fixed my dumb checkstyle errors 21:59:28 <Mierdin> sorry about that 21:59:30 <Mierdin> was dumb 21:59:37 <Mierdin> got vagrant VM up and running now, working through IT 21:59:57 <Madhu> ok. 22:00:29 <Madhu> if u want, we can do hangout and I will show u exactly how I run these tests and code these as well. 22:01:29 <Mierdin> I may need that 22:01:38 <Mierdin> today will be hard, last day of CLUS 22:01:44 <Mierdin> but taking a stab now 22:01:48 <Mierdin> trying to learn :) 22:02:09 <Madhu> sure np. 22:06:03 <Mierdin> getting some errors in the test logs 22:06:06 <Mierdin> io.netty.channel.ConnectTimeoutException: connection timed out: /192.168.111.135:5000 22:06:12 <Mierdin> don't recognize that address 22:06:28 <Mierdin> here's my build string mvn -Dovsdbserver.ipaddress=192.168.50.4 -Dovsdbserver.port=6640 -Pintegrationtest verify 22:07:36 <Madhu> Mierdin: back. 22:07:45 <Madhu> hmm. lemme check. 22:08:26 <Mierdin> btw I verified 50.4 is the actual address, can ping it from my laptop, etc 22:08:35 <Mierdin> oh, and here's how I started docker 22:08:38 <Mierdin> sudo docker run -p 6640:6640 -p 6633:6633 -p 6644:6644 --privileged=true -d -i -t davetucker/docker-ovs:2.1.2 /bin/bash 22:08:56 <Madhu> hmm. the error comes from integration-test.properties file 22:09:08 <Madhu> but i have deleted those files :) 22:09:30 <Madhu> maybe u need a rebase. 22:12:30 <Mierdin> crap 22:12:50 <Madhu> but u shud not be picking up that file anyways. 22:13:11 <Madhu> sorry to ask. u r in topic/schema branch correct :) 22:16:34 <Mierdin> lol yesss :) 22:18:32 <Mierdin> gotta boot for now, will sign on my phone real quick 22:32:53 <networkstatic> signing in from phone to irc is exceptionally nerdy, i like 22:45:21 <ekarlso> networkstatic: feeling naughty ? :p 23:19:06 <tbachman> ekarlso: what r u doing up? 23:25:03 <ekarlso> tbachman: :P 23:25:08 <ekarlso> working on some stuff ;) 23:25:13 <tbachman> crazy man 01:38:52 <odp-gerritbot> Ashwin Raveendran proposed a change to ovsdb: renamed old package to temp and fixes for min/max https://git.opendaylight.org/gerrit/7352 03:38:05 <Mierdin> Hey Madhu 03:38:19 <Madhu> hi Mierdin 03:38:27 <Madhu> guess u r having fun @ clues 03:38:29 <Madhu> clus 03:38:33 <Madhu> saw the pics. 03:38:36 <Madhu> ;) 03:38:51 <Mierdin> Yeah I have a nice bunch of followers lol 03:39:03 <Mierdin> they really like to exploit my embarrasment :) 03:39:57 <Madhu> lol. 03:40:07 <Mierdin> anywho 03:40:12 <Madhu> yes sir. 03:40:16 <Mierdin> I'm really not looking forward to a rebase 03:41:08 <Madhu> :) 03:41:11 <Networkn3rd> He had a great time. 03:41:38 <Madhu> Mierdin: easiest would be to pull a new repo and apply ur patch 03:41:50 <Mierdin> honestly yeah I was going to suggest 03:41:55 <Mierdin> it's not like it's that much code 03:42:04 <Mierdin> yet you wouldn't know it based on how long it's taking me to get this to work :) 03:42:10 <Mierdin> I can do that 03:43:36 <Madhu> Mierdin: IMHO i think this is the best way to learn :) 03:44:08 <Madhu> pro-git is your friend. if you had not reviewed it, would highly recommend it 03:44:15 <Mierdin> as long as I don't annoy you pros too much LOL 03:44:27 <Madhu> small book. but some good info in there 03:44:31 <Madhu> oh god no. 03:45:00 <Madhu> once u have the commit merged in u will be unstoppable :) 03:48:15 <Mierdin> I need to do more practice with things like rebase on a dummy repo 03:48:22 <Mierdin> I have the basics down but the advanced stuff still scares me 03:48:27 <Mierdin> I've rebased before but still....scary 03:49:57 <Madhu> yep, 03:50:05 <Madhu> i have a playground/ directory 03:50:10 <Madhu> specially for git :) 03:50:24 <Madhu> and will recreate scenarios there before monkeying with working code. 03:52:48 <Mierdin> that's a really good practice 03:53:18 <Mierdin> well I am home next week (and over the weekend obviously) and will be working hard to finish this 03:59:31 <Madhu> ok Mierdin don't worry about it. we have many other things to do anyways 03:59:35 <Madhu> and we will concentrate on those 07:14:19 <networkstatic> favorite alias of the day 07:14:40 <networkstatic> alias of13="gsed -i 's/# ovsdb.of.version=1.3/ovsdb.of.version=1.3/g' configuration/config.ini" 07:14:52 <networkstatic> i need a life 07:23:36 <Florian_Zzz> networkstatic, lol 07:24:26 <FlorianOtel> networkstatic, that's just...sad. and funny. 09:40:16 <jordi_cttc> hi, does somebody know how to awake an interface which is down in opendaylight? 09:46:05 <Madhu> networkstatic: i think you should push your aliases to the resources folder :) 09:46:25 <Madhu> i think it will solve world hunger too . lol 09:47:01 <Madhu> jordi_cttc: hi. is this in relation to the port DOWN after stacking ? 09:47:27 <jordi_cttc> yes 09:47:39 <jordi_cttc> madhu: yes 09:47:40 <Madhu> one of the most typical case of port/interface Down is the failure to obtain the Openflow port number. 09:48:00 <Madhu> ODL doesn't influence that at all. that is entirely on the OVS side. 09:48:35 <jordi_cttc> madhu: ok, I imagine ODL is like a transparent stuff between openstack and ovs 09:48:40 <Madhu> jordi_cttc: curious to know your setup that results in a null OF port for theres interfaces 09:49:16 <Madhu> jordi_cttc: kind of. But not so transparent when it comes to Overlay Tunnel management 09:49:29 <jordi_cttc> madhu: I am using the fedora20 VM that Kyle uses in the last presentation 09:49:42 <Madhu> it just doesn't mess up with the libvirt's work on handling the tap ports and addition to br-int 09:49:50 <Madhu> jordi_cttc: okay. 09:49:55 <jordi_cttc> madhu: I made a clon and I am working on with two machines 09:50:14 <Madhu> can u check the ovs logs 09:50:29 <Madhu> under /var/log/openvswitch/ovs-vswitchd.log 09:50:37 <jordi_cttc> madhu: in fact, if you visit http://pastebin.com/7z9J8G8C, you will see what happens after each step I do 09:50:50 <jordi_cttc> madhu: ok, I am going to check the logs 09:50:55 <jordi_cttc> madhu: thanks 09:51:03 <Madhu> sure. this paste bin covers the ODL log. 09:51:09 <Madhu> which is very useful of course :) 09:51:16 <Madhu> but this issue seems to be on the OVS side 09:52:05 <jordi_cttc> madhu: in which computer? controller or compute 09:52:23 <jordi_cttc> madhu: I have them separated and I only count with one compute node 09:52:48 <Madhu> jordi_cttc: it depends on where the problem is seen. 09:52:54 <Madhu> if u look @ this pastebin 09:52:59 <Madhu> tapbd5b518e-31 seems to have a problem 09:53:10 <Madhu> "ERROR o.o.o.n.provider.OF13ProviderManager - Could NOT Identify OF value for port tapbd5b518e-31 on OVS|192.168.56.101:48886" 09:53:27 <Madhu> 56.101 .. which node is that ? 09:53:58 <jordi_cttc> yeah, this is the compute 09:54:35 <jordi_cttc> madhu: I have to mention that sometimes I have the error on the control side and other times in the compute :S 09:55:30 <Madhu> okay. 09:55:49 <Madhu> jordi_cttc: i would just check in both the nodes 09:55:55 <Madhu> and try to spot errors 09:56:30 <jordi_cttc> on both side there are errors 09:56:48 <jordi_cttc> @ the compute: 2014-05-19T09:07:15.164Z|00158|connmgr|INFO|br-int<->tcp:192.168.56.135:6633: sending OFPBRC_BAD_TYPE error reply to OFPST_GROUP request message 09:57:19 <jordi_cttc> @ the controller: 2014-05-19T09:49:41.348Z|00785|connmgr|INFO|Dropped 4 log messages in last 15 seconds (most recently, 15 seconds ago) due to excessive rate 2014-05-19T09:49:41.348Z|00786|connmgr|INFO|br-tun<->tcp:192.168.56.135:6633: sending OFPBRC_BAD_TYPE error reply to OFPST_GROUP request message 09:58:44 <jordi_cttc> right now, the br-int interface (tapXX) of the controller is down 09:59:01 <Madhu> jordi_cttc: can u pls make sure ovs-vswitchd is running ? 09:59:06 <Madhu> ps -aef | grep ovs-vswitch 09:59:24 <jordi_cttc> but sometimes the br-int interface which is down is at the compute node 09:59:58 <jordi_cttc> ps -aef | grep ovs-vswitch root 4852 1 0 08:44 ? 00:00:03 ovs-vswitchd: monitoring pid 4853 (healthy) root 4853 4852 0 08:44 ? 00:00:14 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:err -vfile:info --mlockall --no-chdir --log-file=/var/log/openvswitch/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor odl 8389 988 0 09:59 pt 10:00:04 <jordi_cttc> is working 10:00:11 <Madhu> hmm okay 10:00:30 <Madhu> the tap interface being down has nothing to do with the above error. 10:01:07 <jordi_cttc> I do not why but at the time to make stack.. it start to appear ports at odl and some of them are red 10:01:19 <jordi_cttc> then looking at the odl messages, I saw the error 10:01:35 <jordi_cttc> it is like the openflow staff cannot communicate with the br-X interfaces 10:01:39 <Madhu> can u do "ovs-vsctl list bridge" 10:01:58 <Madhu> also "ovs-vsctl list interface" 10:03:01 <jordi_cttc> the bridge instruction reflects that there are 3 br at the control and 1 at the compute node using all of them OpenFlow13 10:03:32 <Madhu> can u pls paste bin the outputs ? 10:03:43 <jordi_cttc> yes sure 10:04:01 <jordi_cttc> thank you very much madhu, I've been googling the errors but I do not find relevant info 10:06:28 <jordi_cttc> madhu: http://pastebin.com/x69Cceyu 10:06:31 <jordi_cttc> at the controller node 10:06:38 <jordi_cttc> I pass you the link to the compute 10:08:01 <Madhu> jordi_cttc: i do see the gre tunnel created here :) 10:08:17 <Madhu> also the ofport is proper 10:08:37 <Madhu> btw. the error message on the ODL controller doesn't mean it is a persistent error. 10:08:46 <jordi_cttc> ahh okk! 10:08:57 <Madhu> the OVS can throw events in a not so consistent order 10:09:09 <jordi_cttc> regarding the ODL message I though that once was appeared, there was connectivity problems 10:09:14 <Madhu> so ODL might throw an error stating hey i cannot find a corresponding OF port number 10:09:32 <Madhu> but later OVS can allocate it and attach to the Interface 10:09:39 <Madhu> and then ODL will converge and act on them 10:09:44 <jordi_cttc> The log of the compute node: http://pastebin.com/viXQWVvP. I have launched two VM machines and I can ping between them 10:10:18 <jordi_cttc> ahh, ok! so I was looking half the movie ;) 10:10:35 <Madhu> ya the logs look good 10:11:17 <jordi_cttc> madhu: so I am afraid... that I am expecting to see something that already is not integrated ;) 10:11:36 <jordi_cttc> madhu: let me explain what I was planning to test with this setup 10:12:02 <jordi_cttc> both VM's pinging between each other, seems logic to me as they reside in the same compute node. 10:12:30 <jordi_cttc> so what I was planning to do whas to creat another private network, and create another router 10:12:51 <Madhu> L3 is not supported yet 10:12:56 <jordi_cttc> then instante vm's on both private networks and communicate between them 10:13:41 <Madhu> but i have to check if it will work with the existing L3 agent. 10:14:01 <Madhu> jordi_cttc: that is targeted for the Helium release 10:14:02 <Madhu> fyi 10:14:26 <jordi_cttc> by the way, creating a third interface in the default router, I have connectivity between VM's machines placed in different machine 10:14:59 <jordi_cttc> madhu: okkk :), so being a starter, I am trying to run very fast ;) 10:15:16 <jordi_cttc> the Helium release of Openstack? 10:15:24 <jordi_cttc> or odl? 10:16:03 <Madhu> odl 10:16:25 <jordi_cttc> finally, could I ping the VMs from outside? 10:16:32 <jordi_cttc> I mean from the controller? 10:18:00 <jordi_cttc> if I do "route" at the controller, I saw the route to the private network 10.0.0.0 by means of the default router (172.24.4.2) 10:18:25 <jordi_cttc> but when I try, the 172.24.4.1 says that it is a unreachable host 10:18:58 <jordi_cttc> i have ping to 172.24.4.1 10:19:08 <jordi_cttc> maybe the down interface is producing that 10:19:26 <Madhu> u have to use floating ip i think 10:20:04 <jordi_cttc> mmm, if I do that, the stack do not work :s at the controller 10:20:36 <jordi_cttc> with tbachman, we fix local.conf and commenting the floating ip was one point to make stack working 10:20:54 * tbachman wakes up...... 10:21:00 <tbachman> scrolling back now..... 10:21:34 <Madhu> tbachman: is the right person for floating-ip. glad he is up ;) 10:21:43 <tbachman> lol 10:21:44 <jordi_cttc> madhu: jeje 10:21:52 <jordi_cttc> tbachman: hi 10:21:55 <tbachman> We need to test — I can do that now 10:21:58 <tbachman> jordi_cttc: hi! 10:22:17 <tbachman> Madhu: this has been there since the beginning, right? 10:22:23 <tbachman> (i.e. not a recent checkin) 10:22:51 <tbachman> (b/c jordi_cttc is running Fed20, which uses the Hydrogen virtualization release) 10:23:04 <Madhu> tbachman: which recent checkin sir ? 10:23:17 <Madhu> we haven't done anything special for floating. it shud just work :) 10:23:19 <tbachman> Just saying that floating IP support has been there. 10:23:21 <tbachman> Right 10:23:21 <tbachman> k 10:23:42 <Madhu> tbachman: if u remember the trello card (which u volunteered for ;) ) 10:23:53 <Madhu> the floating-ip support is to validate that 10:23:59 <tbachman> Yes — hadn’t caught up with alagalah yet on that, but will just do it now :) 10:26:21 <jordi_cttc> so, why I can ping 172.24.4.1 and not 172.24.4.2? 10:26:26 <jordi_cttc> from the controller node? 10:27:25 <tbachman> jordi_cttc: those are the IPs of your instances? 10:27:30 <tbachman> Or is .1 the router? 10:27:37 <tbachman> (that’s the router…. nm) 10:27:44 <tbachman> jordi_cttc: how are you pinging? 10:27:46 <jordi_cttc> .2 is the router that appears in horizon 10:28:01 <jordi_cttc> ping 172.24.24.2 10:28:06 <tbachman> ah 10:28:26 <jordi_cttc> the vm instances are on the range 10.0.0.0 10:28:43 <jordi_cttc> if I do route at my controller I see: 10.0.0.0 172.24.4.2 255.255.255.0 UG 0 0 0 br-ex 10:29:00 <jordi_cttc> stack creates the route at the controller node 10:29:17 <jordi_cttc> but then I can ping 172.24.4.1 but not 172.24.4.2 10:31:08 <tbachman> jordi_cttc: btw — meant to ask… where are you located? 10:31:11 <tbachman> (time zone) 10:31:16 <tbachman> (if I may ask :) ) 10:32:15 <jordi_cttc> yeah sure, no problem UTC+01:00, Barcelone 10:32:26 <tbachman> Ah, nice! 10:32:35 <tbachman> never been…. but would like to some day :) 10:33:11 <jordi_cttc> really nice place! I recommend you, I can provide you some insights ;) 10:33:18 <tbachman> :) 10:33:20 <jordi_cttc> where are you? if I may ask? 10:33:27 <tbachman> I’m in the U.S., in Maryland 10:33:33 <tbachman> Near the city of Baltimore 10:34:06 <jordi_cttc> uooo, this is what I was thinking, ... so maybe I am disturbing you because it is late at night 10:34:13 <tbachman> lol 10:34:16 <tbachman> no way :) 10:34:18 <tbachman> 6:30 10:34:24 <tbachman> am a morning guy :) 10:34:36 <tbachman> besides… I think if someone is on IRC, they are fair game 10:34:41 <tbachman> people here have odd hours 10:34:46 <tbachman> especially folks like Madhu :) 10:34:52 <tbachman> (3:30 a.m. for him!) 10:35:27 <jordi_cttc> jeje, wow Madhu!! soo then ... again thank you very much!!! 10:36:09 <tbachman> jordi_cttc: the OVSDB channel in particular seems to attract people who don’t sleep ;) 10:36:13 <Madhu> this is the time of the day i can excuse my stupidity ;) 10:36:30 <tbachman> jordi_cttc: Madhu is the ODL rock star 10:36:32 <Madhu> tbachman: its a strategy 10:36:34 <tbachman> don’t listen to him 10:36:47 <tbachman> (on the stupidity part) 10:37:21 <Madhu> lol. i can even claim ignorance :) "did i really say that". i must have been sleep-walking/talking 10:37:31 <Madhu> .. /chatting 10:38:34 <jordi_cttc> lol 10:38:56 <jordi_cttc> I like, you seem crazy ;) 10:41:55 <jordi_cttc> returning to the mestery ;) 10:42:22 <jordi_cttc> why I cannot ping 172.24.4.2 from the controller node 10:42:32 <jordi_cttc> ? 10:44:29 <jordi_cttc> br-ex at the controller node has to have a gre tunnel? .... I think I am starting to get addicted to this ;) 10:54:51 <jordi_cttc> tbachman: maybe I have not created the link point to the exterior 10:55:12 <tbachman> jordi_cttc: sorry — was preoccupied with configuring devstack 10:55:13 <jordi_cttc> tbachman: ovs-vsctl add-port br-ex eth1 10:55:20 <tbachman> so - what is 172.24.4.1? 10:55:23 <jordi_cttc> ooohhh sorry, excuse 10:55:45 <jordi_cttc> I think this is the gateway, which I can ping from the controller node 10:55:52 <tbachman> and 172.24.4.2? 10:56:00 <jordi_cttc> 172.24.4.2 is the IP of the router that appears in horizon 10:56:13 <tbachman> So - the router isnt’ your GW? 10:56:21 <jordi_cttc> it seems so... 10:57:24 <jordi_cttc> in order to do that I was googling and maybe I have missed to configure open-vswitch 10:57:29 <jordi_cttc> tbachman: ovs-vsctl add-port br-ex eth1 10:57:32 <jordi_cttc> with that 10:59:02 <tbachman> jordi_cttc: did you set up the FLOATING_RANGE in local.conf 10:59:03 <tbachman> ? 10:59:09 <jordi_cttc> nooo 10:59:13 <tbachman> ah, k 10:59:14 <jordi_cttc> we removed it 10:59:20 <tbachman> so, it’s the same as where we left it :) 10:59:42 <jordi_cttc> somehow yes 11:00:01 <tbachman> k 11:00:12 <jordi_cttc> according to madhu, the problems I saw at the odl console were not so dramatic... as openvswitch can fix them 11:00:21 <tbachman> let me stack with that — I want to understand just how DS sets up the networking 11:00:35 <tbachman> jordi_cttc: just to make sure I understand — what is it ur trying to do? 11:00:45 <tbachman> Just trying to understand the setup once stacked? 11:01:04 <jordi_cttc> tbachman: I want to test the intregration odl+ostack 11:01:32 <tbachman> So — you are able to stack successfully, and spin up VMs? 11:01:48 <jordi_cttc> tbachman: yes 11:01:57 <tbachman> jordi_cttc: what more do you want to test :) ? 11:02:00 <tbachman> lol 11:02:33 <jordi_cttc> tbachman: jeje, this is what I was trying to check, I thing I am going further that it is already done 11:02:57 <tbachman> not sure — there are other things you can probably play with 11:02:59 <jordi_cttc> tbachman: as both VM's are in the same compute node.. I though it was easy to ping the machines 11:03:08 <tbachman> Oh, wait 11:03:22 <tbachman> you’re trying to ping the VMs from the shell that you ran stack.sh from? 11:03:35 <jordi_cttc> yes, from the controller node 11:03:37 <tbachman> ah 11:03:42 <tbachman> so, there’s another way to do that 11:03:54 <tbachman> you can ping it from the router that’s on their subnet 11:03:59 <tbachman> using “ip netns” commands 11:04:10 <tbachman> jordi_cttc: do you still have the link to that gerrit that I showed you? 11:04:15 <tbachman> the one with the dsconf.sh util? 11:04:21 <jordi_cttc> let me check 11:04:23 <tbachman> If so — there’s another utility in there called pingvm.sh 11:04:27 <tbachman> that will do this for you 11:04:38 <jordi_cttc> tbachman: plenty of utilities, it is like a gold mine 11:04:41 <jordi_cttc> ;) 11:04:44 <tbachman> lol 11:05:04 <tbachman> the shell script isn’t big, so you can peek in to it to see how we’re doing the ping 11:05:15 <jordi_cttc> ok 11:05:21 <tbachman> (credit goes to networkstatic and Madhu for figuring these out — I’m just a packager on this one ;) ) 11:07:27 <jordi_cttc> jeje 11:07:47 <jordi_cttc> tbachman: let me continue explaining my idea... 11:08:51 <jordi_cttc> tbachman: I want to create a second private network and another router to communicate machines in both private networks.. but ... madhu told me that L3 agent is not integrated in odl 11:08:58 <jordi_cttc> so I could not do it 11:09:21 <tbachman> jordi_cttc: I see. I would believe what Madhu says there :) 11:10:51 <jordi_cttc> jeje, me too 11:12:19 <Madhu> oops. lemme explain myself here :) 11:12:27 <Madhu> we haven't tested L3-agent 11:12:37 <Madhu> since we are flooding ARP today 11:12:57 <Madhu> and not removing the L3-agent from the devstack 11:13:18 <Madhu> it could work (as ARP resolution happens behind us) 11:13:43 <Madhu> but.. i have to still check the emac of rule programming for L3-agent 11:14:25 <Madhu> again... l3-agent is something that is flaviof_zzz interest 11:14:36 <Madhu> he is zzz now... but am bet he is interested ;) 11:15:31 <jordi_cttc> jejeje 11:16:34 <jordi_cttc> my computer is close to dye ... too much mambo 11:18:06 <tbachman> lol 11:18:16 <tbachman> jordi_cttc: I’ll see if I can test this out 11:18:22 <tbachman> I still owe Madhu floating IPs :) 11:19:03 <Madhu> tbachman: u r just carrying the sin of others 11:19:06 <Madhu> :) 11:19:11 <tbachman> lol 11:36:50 <tbachman> jordi_cttc: will ping you when I get this resolved 11:37:00 <tbachman> (not sure if it will be today, but hopefully) 12:19:03 <jordi_cttc> tbachman: do not worry tbachman, I will fight it! If we can speak afterwards better! Thanks 16:31:07 <FlorianOtel> . 16:31:30 <tbachman> FlorianOtel: that has a ring of finality to it ;) 16:37:05 <Madhu> dave_tucker: hey buddy 16:37:27 <dave_tucker> hey Madhu 16:37:47 <Madhu> dave_tucker: i think we shud just leave the UT as UT :-) 16:38:01 <Madhu> just saw Ashwin's comments and I agree with him on that (as u know already) 16:38:06 <dave_tucker> Madhu: it's not a UT :) but point taken 16:38:13 <Madhu> lol. yes it is not 16:38:16 <Madhu> but it is not IT as well 16:38:19 <Madhu> but anyways. 16:38:26 <Madhu> regarding the failure retry 16:38:38 <Madhu> though the retry will work in this case. 16:38:55 <Madhu> how about the IT with docker instances contending for the same port on the host machine ? 16:39:00 <Madhu> won't that cause problems too > 16:39:22 <dave_tucker> no sir... as there is only a single slot open on that jenkins node 16:39:34 <dave_tucker> or at least there should be only one ;) 16:40:08 <Madhu> hmm. 16:40:09 <Madhu> then how come the UT fails ? 16:40:15 <Madhu> because we cannot have multiple parallel jobs. isn't it 16:40:46 <dave_tucker> correction. ovs-verify has one slot on the jenkins slave. the normal verify job has 2 slots 16:41:08 <Madhu> maybe we shud just change the normal verify to 1 slot as well 16:41:12 <Madhu> problem solved ;) 16:41:20 <Madhu> anyways we have to do that for IT 16:44:03 <dave_tucker> Madhu: lol. but then we have longer build queues 16:44:20 <Madhu> dave_tucker: thats okay. because we have this problem in IT anyways 16:44:46 <Madhu> and we have to enable IT to the regular Verify and merge jobs 16:45:01 <Madhu> don't want to change the code because of setup environment issues :) 16:46:08 <dave_tucker> up to you Madhu... as long as we have tests and they run I don't mind 16:46:18 <dave_tucker> but would prefer that we could fix the IT's to run in parallel 16:46:53 <Madhu> dave_tucker: its a valid point on the longer queues 16:47:17 <dave_tucker> Madhu: my worry is when we move to dyanamic slaves, that we could have conflicts with other projects 16:47:20 <Madhu> maybe we can first get the IT environment fixed to have it a capability to run in parallel 16:47:38 <Madhu> that won't happen :) 16:47:49 <Madhu> tykeal is awesome in giving us dedicated setup 16:48:12 <dave_tucker> Madhu: we should catch up with tykeal when he's back 16:48:18 <Madhu> yep. 16:48:36 <Madhu> dave_tucker: for the moment, lets freeze that gerrit and come back to it with a plan 16:48:38 <dave_tucker> iirc he was against us having special treatment 16:48:45 <Madhu> not us :) 16:48:54 <Madhu> it is for most of the integration tests today 16:49:01 <Madhu> and I don't see any other way actually 16:49:17 <dave_tucker> ok Madhu: we just need to note that jenkins builds marked unstable may not actually be unstable for now 16:49:18 <Madhu> well. theoretically yes. but practicaly it is a nightmare to share resources that way 16:49:40 <Madhu> shall we make the verify job for 1 slot ? 16:50:09 <Madhu> and am inclined to enable the ova-verify job for the schema branch triggers. 16:50:14 <Madhu> what do u think ? 16:50:18 <dave_tucker> Madhu: YES! 16:50:40 <dave_tucker> I was going to do that, but we still have a bug that is causing intermittent IT failure 16:50:41 <Madhu> cool. 16:50:47 <Madhu> btw. mutate works 16:50:54 <dave_tucker> :) 16:51:13 <Madhu> and i will send the patch to Matt to take a look 16:52:20 <dave_tucker> Madhu: re jenkins, we could uncheck "Execute concurrent builds if necessary" 16:52:32 <dave_tucker> but we'll have to do this for both ovsdb-merge and ovsdb-verify 16:52:34 <Madhu> yep. 16:52:40 <Madhu> np. lets do it 16:52:45 <Madhu> ours is not high volume anyways 16:52:46 <dave_tucker> ok sir. I'm logged in 16:52:49 <dave_tucker> i'll do it now 16:52:50 <Madhu> thanks buddy 16:53:20 <dave_tucker> Madhu: actually sir it won't work 16:53:27 <Madhu> :( 16:53:35 <dave_tucker> merge and verify could be run in parallel 16:53:51 <Madhu> hmm... 16:53:56 <dave_tucker> need to open a helpdesk ticket for the jenkins master to only have one open build slot 16:54:13 <dave_tucker> that would be best I think 16:54:21 <Madhu> yes. sir. good point 16:57:41 <rrsantos> hello all! does any of you tried to set up openflow rules in a multihomed node with openvswitch, but at the same time, as an endpoint (receiving and sending application traffic)? 16:58:27 <dave_tucker> Madhu: can you take a look at this one today? would like to get this merged ;) https://git.opendaylight.org/gerrit/#/c/7122/ 16:58:51 <Madhu> yes dave_tucker certainly 16:59:29 <dave_tucker> thanks Madhu. have abandoned the UT/IT change but have take a todo to refactor that test to look more like a conventional UT 16:59:36 <dave_tucker> will make my code OCD happy :) 17:00:24 <dave_tucker> gtg put myles to bed. ttfn 17:13:26 <Madhu> rrsantos: hi 17:13:42 <rrsantos> Madhu: hello 17:13:54 <Madhu> i have done something like that 17:14:03 <Madhu> is there a question on that rrsantos ? 17:14:48 <rrsantos> well yes, how did you made it? I was considering multiple tap interfaces 17:31:12 <rrsantos> Madhu: I was considering something like this http://i.imgur.com/XQgpHnk.png 17:34:14 <Madhu> rrsantos: yes. I did this exactly 17:34:32 <Madhu> and infact our devstack setup has something like this already 17:34:41 <Madhu> all the VMs launched by Nova are Tap interfaces 17:35:01 <Madhu> and the middle layer of OVS are hosted on another node. 17:36:27 <rrsantos> ahhh that's great, at least I already have that thing figured out. I think my current issue right now is about routing 17:36:37 <rrsantos> and setting up matching flows 17:37:09 <rrsantos> I'll have to go now, but meanwhile I'll dig up a bit more on that subject. Thank you Madhu :) 17:39:30 <networkstatic> ugh, been so heads down there has been an infomercial in the background for the past 12 hours on repeat 17:39:53 <networkstatic> i got good tip on aging secrets now 17:40:34 <mlemay> LOL 17:40:46 <networkstatic> hehe hey mlemay 17:41:39 <mlemay> hey Brent.. so got the flow stuff working? 17:42:30 <networkstatic> good man, done today 17:42:51 <networkstatic> far too many projects on this one 17:43:01 <networkstatic> like 6 or something 17:43:42 <networkstatic> we here in ovsdb like to turtore ourselves using as many OSS projects as possible 17:44:02 <networkstatic> like a mini interop project 18:40:49 <odp-gerritbot> Brent Salisbury proposed a change to ovsdb: Adding neutron api v2.0 security group/rules handlers to the Neutron integration in OVSDB plugin. Depends on pending neutron security apis being reviewed. https://git.opendaylight.org/gerrit/7364 18:59:25 <odp-gerritbot> Brent Salisbury proposed a change to ovsdb: Adding neutron api v2.0 security group/rules handlers to the Neutron integration in OVSDB plugin. Depends on pending neutron security apis being reviewed. https://git.opendaylight.org/gerrit/7366 19:00:41 <FlorianOtel> . 20:45:46 <Madhu> dfarrell07: hi. can u share the shared calendar ? 20:47:33 <dfarrell07> Madhu: Yeah, one sec, digging up the link 20:48:00 <dfarrell07> http://goo.gl/mxwjTb 20:48:18 <dfarrell07> You can use the button at the bottom right to subscribe to the cal 20:56:08 <dfarrell07> Madhu: You ment share it here, right? Or more widly? 20:56:17 <Madhu> dfarrell07: thanks sir 20:56:53 <dfarrell07> Madhu: np :) 21:00:08 <Mierdin> thx dfarrell07 that calendar is handy 21:00:17 <Mierdin> official? 21:00:42 <dfarrell07> Mierdin: Thanks! No, the wiki should be considered the offical source. 21:00:52 <dfarrell07> It does contain all the data from all the wikis though 21:01:34 <dfarrell07> So unoffically, it's equivalently offical ;) 21:02:03 <dfarrell07> Mierdin: https://wiki.opendaylight.org/view/Meetings 21:03:55 <Mierdin> dfarrell07: rock on! 22:07:00 <networkstatic> the miniature keyboard vacuum never really took off did it 22:09:14 <padma_> I am looking for REST API equivalent of "addTunnel <Node> <Bridge> <Port> <tunnel-type> <remote-ip>" to automate my testing 22:09:34 <networkstatic> standby padma_ will get u the link 22:09:51 <padma_> thank you 22:10:19 <padma_> one more help; which module has implementation of this command in ovsdb? 22:11:45 <networkstatic> here you go buddy 22:11:46 <networkstatic> https://git.opendaylight.org/gerrit/gitweb?p=ovsdb.git;a=blob;f=resources/commons/OVSDB_Northbound_APIs.json.postman_collection;h=97a09e4390ce468e3a27eae46873e125a1d2cf16;hb=HEAD 22:12:31 <networkstatic> ovsdb/northbound/ovsdb bundle and ConfigurationService in ovsdb/ovsdb 22:13:30 <networkstatic> apologies calling you buddy btw in case ur a female. the name didnt register with me 22:13:50 <padma_> the postman collection i didn't see POST for addTunnel. I saw bridge, port interface add 22:13:54 <networkstatic> ah 22:14:18 <Madhu> padma: the postman collection has all the generic info 22:14:26 <networkstatic> lemme see if i have an example in my local collection 22:14:27 <Madhu> and you have to construct it for whatever u need. 22:14:32 <Madhu> including the addTunnel 22:14:45 <networkstatic> good point could be handy 22:16:21 <padma_> madhu that's exactly what I am looking for 22:18:19 <Madhu> padma_: http://openvswitch.org/ovs-vswitchd.conf.db.5.pdf is the actual schema that has all the fields exposed by ovsdb 22:18:37 <Madhu> and please use the create Port and create Interface REST-API from that post-man collection 22:18:56 <Madhu> and fill in any data that you want that helps with creating the tunnel base on the schema above 22:19:21 <Madhu> padma_: we can create a specific example of creating a tunnel in post-man 22:19:46 <padma_> yes that would be helpful 22:19:48 <Madhu> but it would be a good exercise for you to use the above schema and Postman collections and will help you manage the entire switch 22:20:10 <networkstatic> thanks madhu, tty everyone in a bit, if i dont have a great weekend. 22:21:50 <Madhu> padma_ I am middle of something now. pls drop me an email and I will give u the exact postman script that takes care of this. 22:21:56 <Madhu> mavenugo@gmail.com 22:22:42 <padma_> thanks Madhu 21:59:58 <dave_tucker> networkstatic: you about sir? 22:02:56 <dave_tucker> ping me if you want help with the neutron stuff bro... can help get the services registered in osgi ;) 22:07:56 <networkstatic> cool thanks, was bouncing the APIs off madhu on updates then will put up the patches for review. 22:08:33 <networkstatic> the ovsdb part wont build currently as it doesnt have the services to subscribe from in controller yet 22:09:33 <networkstatic> just put the patch up as thats the NB ovsdb knowing it would fail build until dependancies get up there 22:10:03 <networkstatic> dave_tucker_zzz ah u slept 20:36:27 <networkstatic> gawd, i just slept 9 hours 20:36:35 <tbachman> networkstatic: nice! 20:36:44 <tbachman> when’s the last time *that* happened? 20:36:47 <networkstatic> wtf, im like a vampire now 20:36:51 <tbachman> lol 20:37:01 <networkstatic> i need to start wearing mascarra 20:37:08 <tbachman> goth 20:37:08 <networkstatic> get weird with it 20:37:13 <tbachman> own it 20:37:16 <networkstatic> yes 20:37:33 <networkstatic> so we are off tomorrow 20:37:35 <tbachman> were u running an all-nighter before? 20:37:41 <tbachman> networkstatic: we’re supposedly off now 20:37:46 <networkstatic> every night :) 20:37:50 <tbachman> seems like we’re not so good with these things 20:37:51 <tbachman> lol 20:37:55 <networkstatic> no not at all 20:38:16 <networkstatic> i dont think ive taken a "day off" since ODL became real 20:38:22 <networkstatic> wtf would i do? 20:38:24 <tbachman> lol 20:38:25 <networkstatic> go play golf? 20:38:29 <tbachman> save kitties 20:38:41 <networkstatic> do that on laptop breaks already 20:38:42 <networkstatic> lol 20:38:43 <tbachman> build the NSA roomba 20:38:50 <networkstatic> see now that 20:38:56 <networkstatic> is what we should be doing 20:39:10 <networkstatic> Danial is on the robot team or some shit 20:39:14 <networkstatic> our intern 20:39:24 <networkstatic> we should have him sweat shop it out 20:39:29 <tbachman> folks think that the innocent, cute robot on the floor is just cleaning it up 20:39:32 <networkstatic> make money on his labor 20:39:38 <tbachman> really a networked set of super-sensors 20:39:43 <tbachman> for gvmt 20:39:45 <networkstatic> child labor 20:39:47 <tbachman> lol 20:39:51 <networkstatic> yes 20:40:41 <networkstatic> FlorianOtel 20:40:45 <tbachman> networkstatic: no TWS on Monday, right? 20:40:55 <networkstatic> ah u r the best 20:40:58 <networkstatic> thank u 20:41:01 <networkstatic> i send an email now 20:42:51 <networkstatic> man, what a team 07:54:43 <FlorianOtel_> . 13:28:18 <jordi_cttc> hi guys 15:47:06 <Madhu> padma: just replied to ur email 09:36:55 <jordi_cttc> hi guys 09:37:24 <jordi_cttc> has someone connected a vm to the internet with the integration odl+openstack (fedora20)? 13:27:46 <odp-gerritbot> A change was merged to ovsdb: Sample OpenStack/Neutron v2.0 Postman API Collection https://git.opendaylight.org/gerrit/7409 15:49:44 <odp-gerritbot> A change was merged to ovsdb: Adding neutron api v2.0 security group/rules handlers https://git.opendaylight.org/gerrit/7366 18:07:01 <jimooo> networkstatic: A week or two ago you were irritated with the openflow folks because they were breaking things in the code. I'm thinking about updating to the latest builds. Have things stabilized? 18:42:54 <Madhu> Mierdin: hi 18:42:57 <Madhu> networkstatic: hey buddy. 18:50:47 <tbachman> Madhu: are we doing WebEx or hangout today? 18:51:01 <Madhu> webex i think :) 18:51:05 <tbachman> k 19:07:27 <Madhu> hi LuisGomez r u joining us ? 19:13:30 <CASP3R> https://wiki.opendaylight.org/view/OpenDaylight_OpenFlow_Plugin:Test_Provider 19:13:35 <CASP3R> is that the link? 19:54:18 <Madhu> Mierdin: hi u r up next on the call sir :) 21:03:29 <dfarrell07> Madhu, networkstatic: You guys are cool with having the *_integration jobs Ed mentioned, right? I'd gladly take them on. Just didn't hear a clear "ACK, yeah that's a thing we want, someone work on it" 21:04:16 <Madhu> dfarrell07: let me check out the Integration projects and check what is being done there 21:04:27 <dfarrell07> Madhu: kk 21:04:33 <Madhu> and i would prefer dave_tucker_zzz around to make that call as well. 21:04:41 <dfarrell07> Sure, makes sense 21:04:48 <Madhu> adding an integration job with a trigger is not a big deal. 21:05:00 <Madhu> just that I want to make sure it all falls in line with the work that we are already taking up 21:05:09 <dfarrell07> Right, I just took a look and I think I already see clearly how to do it 21:05:15 <dfarrell07> kk 02:11:38 <Madhu> Mierdin: hi matt 02:11:51 <Madhu> got your email and yes. -p indicates the port forwarding. 02:12:02 <Madhu> the port forwarding instructions are incorrect 02:28:38 <Madhu> Mierdin: use just -p 6641:6640 02:45:27 <Madhu> Mierdin1: just replied to ur email 09:08:20 <phate06> a stupid question about openflow 09:08:38 <phate06> two rules identical, one has a field more in the matching 09:09:11 <phate06> will it be taken the most specific one or one will overwrite the other? 09:23:08 <networkstatic> highest priority wins 09:24:07 <networkstatic> sorting based on priority, so top to bottom , first match is the end of the table 09:24:45 <networkstatic> one field at p=50000 would be elected over a 12-tuple at p=49000 09:24:59 <networkstatic> that make sense? 09:42:26 <phate06> and if they have the same priority? 09:43:01 <phate06> but one has a field more, for example one has an ingress port specified 09:43:16 <phate06> on a switch with 3 ports 09:43:25 <phate06> if a flow comes from port 1 09:43:48 <phate06> would it be taken the flow rule specifing ingress port=1 over the one which doesn't specify an ingress port at all? 10:06:06 <lori> phate06: if the priority is not the same, the result is not very deterministic 10:06:37 <lori> phate06: AFAIK, the first rule will be used for the match 10:07:09 <lori> phate06: what I'm sure about is that the number and type of match fields does not matter 10:07:22 <lori> phate06: no such thing as a more specific match 10:51:52 <phate06> ok too bad... thanks for the answer:) 20:10:34 <networkstatic> anyone know how to disable IPv6 in local.conf on the compute node? 20:15:13 <Mierdin1> networkstatic: for SHAME 20:15:23 <networkstatic> lolz 20:15:34 <networkstatic> dont u dare tell ed 20:15:36 <networkstatic> :) 20:17:01 <Mierdin1> how much is my silence worth to you? :) 20:32:11 <Mierdin1> Madhu: would love to formally document the vagrant stuff for you guys if that's cool 20:32:21 <Mierdin1> there's a brief mention on the main page but would love a full walkthrough 20:32:26 <Madhu> Mierdin1: pls do sir :) 20:32:35 <Mierdin1> how do I go about getting perms to do so? 20:32:41 <Madhu> Mierdin1: we can't have enough documentation 20:32:53 <Madhu> perms are all open for all in the wiki :) 20:33:08 <Madhu> you can go ahead and add a page under ovsdb and start docing:) 20:34:57 <networkstatic> +Cats 20:35:09 <networkstatic> im tired of +1s, too impersonal 20:35:16 <networkstatic> im not a machine 20:36:15 <Madhu> networkstatic: purr purr 20:36:18 <Madhu> tired of LOLs ;) 20:36:28 <networkstatic> LOL 20:36:31 <networkstatic> PURRR 20:36:48 <networkstatic> IM TIRED OF LOWER CASE LETTERS TOO 20:36:54 <networkstatic> FROM HERE ON OUT 20:36:57 <networkstatic> IM SHOUTING 20:37:15 <Madhu> your PURRR is like ROAR 20:37:18 * Madhu scared 20:37:18 <networkstatic> hahaha 20:37:56 <Madhu> and this will be tweeted soon : "networkstatic: anyone know how to disable IPv6 in local.conf on the compute node?" 20:38:01 <networkstatic> Time to get serious https://www.dropbox.com/s/i4qs9uy3xt11yw8/Screenshot%202014-05-28%2016.37.58.png 20:38:02 <Madhu> maybe cc Ed 20:38:04 <networkstatic> NOOOOOOO 20:38:12 <networkstatic> lol 20:38:22 <Madhu> :) 20:38:33 <Madhu> serious business. love it. 20:38:34 <networkstatic> the v6 harrasment is like a dose of my SDN twitter trolling 20:38:53 <networkstatic> i dont remember anything about v6 20:38:59 <networkstatic> other then it looks like a mac address 21:10:44 <flaviof> networkstatic: in fedora, you can do it by changing grub: http://www.server-world.info/en/note?os=Fedora_19&p=initial_conf&f=3 21:11:10 <flaviof> networkstatic: https://www.dropbox.com/s/gwtxyp5q81v1abx/Screenshot%202014-05-28%2017.11.07.png 21:11:46 <flaviof> networkstatic: I like the "serious" pic :) 21:21:24 <networkstatic> ah nice thanks flaviof didnt think about just disabling :) 21:46:28 <jimooo99> I'm trying to read the observation ID from the IPFIX object. It's defined to return OvsDBSet<Integer> but seems to be actually retirning OvsDBSet<BigInteger> 21:57:23 <Madhu> jimooo99: i don't know of anyone testing ipfix table :) other than you. 21:57:33 <Madhu> so please file a bug whenever u find 1 21:57:42 <Madhu> and 1 of us (including you :) ) can fix that 22:04:41 <jimooo99> the latest builds seem really broken. Can't run on windows at all and now it just connects and disconnects repeatedly in linux. So many exceptions that my own logging gets lost. 01:33:41 <Mierdin1> I presume everyone's planning on attending this? http://events.linuxfoundation.org/events/opendaylight-developer-design-forum 01:33:49 <Mierdin1> About to request invite before I forget 02:37:31 <networkstatic> Mierdin roger buddy 02:37:38 <networkstatic> thats the next big one 03:29:16 <Mierdin> networkstatic: cool, sent 04:33:07 <networkstatic> Mierdin excellent 01:56:39 <snowzach> Hi folks 01:57:04 <snowzach> Anyone willing to help a guy out trying to get opendaylight+openstack to work? 01:58:16 <snowzach> @Mierdin, I cannot think of a reason why you could not have a compute node in diff DC assuming your network plugin worked. 01:59:01 <snowzach> The local and vlan neutron plugins assume you are plugged into the same switch 02:17:07 <snowzach> I've got a 3 compute nodes and what I call my control/network node plus opendaylight 02:17:35 <snowzach> Got ML2 configured with opendaylight driver, all my ovs instances pointed to opendaylight 02:17:58 <snowzach> looks like opendaylight sets up VXLan tunnels between all the OVS instances 02:18:22 <snowzach> When I try to boot a VM, I get: {u'message': u'Unexpected vif_type=binding_failed', u'code': 500, u'details': u' File "/usr/lib/python2.7/site-packages/nova/compute/manager.py", line 258, in decorated_function 10:46:13 <pix_> hello world 19:19:29 <tykeal> #endmeeting