Tuesday, 2016-10-04

*** amotoki has quit IRC00:00
*** amotoki has joined #tacker00:01
*** amotoki has quit IRC00:05
*** amotoki has joined #tacker00:13
*** s3wong has quit IRC00:54
*** amotoki has quit IRC00:59
*** amotoki has joined #tacker01:00
*** prashantD has quit IRC01:00
*** amotoki has quit IRC01:04
*** amotoki has joined #tacker01:08
*** amotoki has quit IRC01:24
*** amotoki has joined #tacker01:25
*** amotoki has quit IRC01:29
*** yifei has joined #tacker01:49
*** uck has joined #tacker01:59
*** amotoki has joined #tacker01:59
*** yifei has quit IRC02:16
*** uck has quit IRC02:25
*** uck has joined #tacker02:25
*** vishnoianil has quit IRC02:29
*** uck has quit IRC02:30
*** yifei has joined #tacker02:44
*** amotoki_ has joined #tacker02:55
*** amotoki has quit IRC02:58
*** links has joined #tacker03:10
*** yifei has quit IRC03:16
*** prashantD has joined #tacker03:23
*** yifei has joined #tacker03:25
*** yifei has quit IRC03:34
*** vishwanathj_ has quit IRC04:14
*** amotoki has joined #tacker04:18
*** amotoki_ has quit IRC04:21
*** lamt has quit IRC04:31
*** KanagarajM has joined #tacker04:33
*** prashantD has quit IRC04:56
*** diga_ has joined #tacker04:57
diga_sridhar_ram: Hi04:57
*** Ravikiran_K has joined #tacker05:05
*** neel has joined #tacker05:10
*** vishnoianil has joined #tacker05:23
*** yifei has joined #tacker05:30
*** yifei has quit IRC05:34
*** neel has quit IRC05:44
*** manikanta_tadi has quit IRC05:45
*** manikanta_tadi has joined #tacker05:45
*** neel has joined #tacker05:55
*** yifei has joined #tacker06:13
*** KanagarajM has quit IRC06:15
*** yifei has quit IRC06:22
*** neel has quit IRC06:26
openstackgerritgecong proposed openstack/tacker: Replace LOG.warn with LOG.warning  https://review.openstack.org/38143506:29
*** Ravikiran_K has quit IRC06:29
openstackgerritSharat Sharma proposed openstack/python-tackerclient: Changed the link to home-page  https://review.openstack.org/38143606:32
*** amotoki has quit IRC06:43
*** amotoki has joined #tacker06:49
*** amotoki_ has joined #tacker07:24
*** amotoki has quit IRC07:27
*** xuhaiwei_ has joined #tacker07:36
xuhaiwei_sridhar_ram: ping07:37
sridhar_ramxuhaiwei_: hi07:37
xuhaiwei_I have added a topic in Barcelona summit session page https://etherpad.openstack.org/p/tacker-ocata-summit07:38
xuhaiwei_can we have a discussion about it? senlin integration with tacker, sridhar_ram07:38
sridhar_ramxuhaiwei_: sure, are you going to be in the summit ?07:39
xuhaiwei_sridhar_ram, yes07:39
sridhar_ramxuhaiwei_: good.. i'll move this item up into the main schedule..07:39
xuhaiwei_sridhar_ram, thanks a lot07:40
sridhar_ramxuhaiwei_: btw, quick question on senlin usage for tacker..07:40
xuhaiwei_sridhar_ram, sure07:40
xuhaiwei_sridhar_ram, senlin can help mentaince VNF auto-scaling, HA and so on07:41
sridhar_ramxuhaiwei_: beyond what is written in the above BP, i think senlin can also be used for auto-scaling.. is that a different use case of senlin ?07:41
sridhar_ramxuhaiwei_: but the scope of your BP is only to address HA ?07:42
xuhaiwei_sridhar_ram, Senlin provides a service of clustering management, including auto-scaling07:42
sridhar_ramxuhaiwei_: currently tacker is using heat auto-scaling.. can heat auto-scaling co-exist with VDUs marked for HA using  senlin ?07:43
xuhaiwei_sridhar_ram, yes, the BP is created by other developer, I will make an introduction fo Senlin in summit07:44
*** n-harada has joined #tacker07:44
sridhar_ramxuhaiwei_: sounds good.. that will be useful.. btw, please keep sripriya in the loop07:44
*** vishnoianil has quit IRC07:45
xuhaiwei_sridhar_ram, yes, senlin can work as a intra_driver instead of heat, that means tacker will talk to senlin first and then to heat, also senlin can work under heat, as a heat resource, in that case it will be tacker -> heat -> senlin -> heat07:46
sridhar_ramxuhaiwei_: tacker -> heat auto-scaling -> senlin make sense to me..07:47
sridhar_ramxuhaiwei_: however, heat -> senlin -> heat .. ? that doesn't make sense..07:48
sridhar_ramxuhaiwei_: anyways, these are good topics for a F2F summit design session07:48
sridhar_ramxuhaiwei_: please prepare a short 3 - 5 slides presentation for tacker design summit session in senlin07:49
xuhaiwei_sridhar_ram: if use senlin, you don't need to use heat auto-scaling, senlin uses heat to do orchestration, for auto-scaling, senlin uses its own way to do it07:49
sridhar_ramxuhaiwei_: okay, i think i understand..07:49
xuhaiwei_sridhar_ram, sure, I will prepare for it07:49
xuhaiwei_sridhar_ram: :)07:51
*** vishnoianil has joined #tacker07:53
*** yifei has joined #tacker07:59
*** Ravikiran_K has joined #tacker08:09
*** diga_ has quit IRC08:12
*** amotoki_ has quit IRC08:18
*** diga_ has joined #tacker08:18
*** amotoki has joined #tacker08:21
*** yifei has quit IRC08:22
*** n-harada has quit IRC08:31
*** yifei has joined #tacker08:49
*** yifei has quit IRC08:54
*** dkushwaha has quit IRC09:29
*** amotoki has quit IRC09:51
*** prashantD has joined #tacker09:54
*** prashantD has quit IRC09:59
*** amotoki has joined #tacker10:07
*** amotoki has quit IRC10:39
*** amotoki has joined #tacker10:40
*** neel has joined #tacker10:48
*** uck has joined #tacker11:17
*** KanagarajM has joined #tacker11:24
*** bobh has joined #tacker11:25
*** yifei has joined #tacker11:26
*** yifei has quit IRC11:30
*** janki has joined #tacker11:33
*** uck has quit IRC11:45
*** bobh has quit IRC11:54
*** amotoki has quit IRC11:57
*** KanagarajM has quit IRC12:09
*** amotoki has joined #tacker12:12
*** KanagarajM has joined #tacker12:13
*** neel has quit IRC12:15
*** Ravikiran_K has quit IRC12:15
*** KanagarajM has quit IRC12:23
*** diga_ has quit IRC12:27
*** amotoki has quit IRC12:49
*** amotoki has joined #tacker12:54
*** dkushwaha has joined #tacker13:02
*** janki has quit IRC13:09
*** Vijayendra has quit IRC13:18
trozetsridhar_ram: pong13:20
*** amotoki has quit IRC13:38
*** yifei has joined #tacker13:44
*** bobmel has joined #tacker13:44
*** yifei has quit IRC13:49
*** links has quit IRC13:49
*** Vijayendra has joined #tacker13:55
-openstackstatus- NOTICE: The Gerrit service on review.openstack.org is being restarted to address performance degradation and should return momentarily14:02
*** amotoki has joined #tacker14:17
*** vishwanathj has joined #tacker14:53
trozetsridhar_ram: around?14:54
*** bobh has joined #tacker14:58
*** bobh has quit IRC15:16
*** vishnoianil has quit IRC15:16
*** sripriya has joined #tacker15:23
*** vishnoianil has joined #tacker15:28
*** prashantD has joined #tacker15:40
*** diga_ has joined #tacker15:59
*** lamt has joined #tacker16:01
*** amotoki has quit IRC16:03
sridhar_ramFolks - weekly meeting time!16:03
sridhar_ramjoin #openstack-meeting16:03
*** bobh has joined #tacker16:05
*** links has joined #tacker16:17
*** links has quit IRC16:18
*** openstackgerrit has quit IRC16:26
*** openstackgerrit has joined #tacker16:27
*** openstackgerrit has quit IRC16:28
trozetdiga, diga_, sridhar_ram: did you compile the openvswitch packages from source? or did you download the debs somehwere?16:29
*** openstackgerrit has joined #tacker16:29
sridhar_ramtrozet: didn't compile from src, used the base ovs 2.5 that came w/ Ubuntu 16.0416:30
*** openstackgerrit has quit IRC16:30
trozetsridhar_ram: heh my 16.04 doesnt have it, I built from source first time, but somethign was broken16:30
trozetsridhar_ram: you were just using 2.5.0?16:30
*** openstackgerrit has joined #tacker16:31
diga_trozet:  I have just installed ubuntu16.04 with tacker but will install ovs2.516:32
trozetdiga_ so you already had ovs2.5 on your ubuntu, or did you get it from somewhere?16:32
diga_trozet: I installed openvswitch-ipsec16:35
diga_which is 2.5.016:35
diga_trozet: hope its correct installable ?16:36
trozetdiga_: i dont think you need ipsec16:36
diga_trozet: okay16:36
trozetdiga_: openvswitch-switch_2.5.0-0 openvswitch-common_2.5.016:37
diga_trozet: okay, will install it manually,16:38
*** Ravikiran_K has joined #tacker16:41
sridhar_ramtrozet: I just pulled ovs using 'apt-get install openvswitch-vswitch"17:01
diga_sridhar_ram: sripriya: Hi, did you get time to look at htis spec - https://review.openstack.org/#/c/368511/17:01
trozetsridhar_ram: ok, i found the a deb online and downloaded, looks like things are stacking now17:01
sridhar_ramtrozet: cool17:02
sridhar_ramdiga_: not yet fully..17:02
*** dkushwaha has quit IRC17:03
diga_sridhar_ram: please review it once you get time, eager to start coding :)17:03
diga_trozet: can you share the link, so that I can download the same17:04
sridhar_ramdiga_: one general concern i've is .. API / conversion to pecan is a bigger undertaking..17:04
sripriyadiga_: not yet, did you plan to cover a substantial report of bringing in the pecan framework, i’m looking around similar lines of pecan spec in neutron (since our api framework was derived from neutron originally) http://specs.openstack.org/openstack/neutron-specs/specs/kilo/pecan-switch.html17:04
diga_sripriya: okay17:05
sridhar_ramdiga_: sripriya: also my understanding is it will most likely force us to do a v2 API bump..17:06
*** dkushwaha has joined #tacker17:07
sridhar_ramdiga_: again, i'd suggest you to take up smaller RFEs first.. familiarize with the tacker code base first before taking this up..17:07
diga_sridhar_ram: I think you know or not but I am the only person who has written first WSGI+PECAN based implementation for magnum17:08
sripriyasridhar_ram: i am not sure there is any change on the API signature itself for the resources, but good to think about..17:08
sridhar_ramdiga_: ah, i didn't know that.. in fact u shd guide us also in the magnum exploration!17:09
diga_sridhar_ram: :) yeah! sure17:09
sridhar_ramsripriya: I'll be happy if switching to pecan *doesn't* force us a version bump17:10
diga_sridhar_ram: I have implemented complete docker backend in the magnum along with COE part with apmelton17:10
sridhar_ramdiga_: question, during magnum pecan work... did you use a separate branch, developed there and merged to master ?17:10
diga_sridhar_ram: I have created seperate project on my github repo & then merged it to magnu17:11
diga_magnum*17:11
sridhar_ramdiga_: super cool.. can u share the link ?17:11
diga_sridhar_ram: https://github.com/digambar15/magnum17:14
diga_sridhar_ram: this is very first implementation17:15
sridhar_ramdiga_: did magnum start out with another wsgi component and then switched to Pecan ? Or is Pecan is the first and only WSGI framework introduced in magnum ?17:15
diga_sridhar_ram: I think ceilometer used firstly PECAN+WSME17:16
diga_sridhar_ram: but when we started the implementation, it was at very early stage in ceilometer, so we built it from scratch17:16
sridhar_ramdiga_: I see...17:17
diga_sridhar_ram: now many openstack project are using WSME+PECAN, neutron, ceilometer, watcher etc17:17
diga_sridhar_ram: but talked with many people, I am working with nova scheduler team also but it seems everyone prefers PECAN over falcon17:18
diga_sridhar_ram: though its big impact but its not that big like neutron :)17:19
sridhar_ramdiga_: I see..then i'd encourage you to start some prototype working in a branch off tacker17:19
diga_Thanks sridhar_ram17:19
sridhar_ramdiga_: what are some of the risks in this venture ?17:20
diga_sridhar_ram: I think whole API level changes will be required along with model path. We need to restructure everything related to API according to PECAN approach17:21
sridhar_ramdiga_: I'm wondering if we should introduce pecan as a configurable option first and then switch it to default and remove the "older" framework in release N+117:21
diga_sridhar_ram: endpoint will not change17:21
diga_sridhar_ram: that can be the good option17:22
diga_sridhar_ram: let other people work on existing framework17:22
diga_sridhar_ram: what is the timeline for N+1 ?17:22
*** bobh has quit IRC17:22
diga_sridhar_ram: so that I can make sure it will be ready & tested before merge to N+117:23
*** bobh has joined #tacker17:23
sridhar_ramdiga_: I meant, introduce this in Ocata as an optional framework.. but switch to it being the default in next P release17:23
diga_sridhar_ram: Yeah, okay17:23
sripriyasridhar_ram: diga_: these are the questions i’m looking for an answer in spec, with tacker’s current state, what are the specifc problems that are addressed by pecan, for example: do i still need to maintain the resource attributes as dictionary inside each extension? what are the advantages the new changes will bring in? how are the validations of attributes done? etc.17:24
trozetsridhar_ram: do you have your local.conf handy.  i stacked OK but cannot reach neutron 40417:24
diga_sripriya: According to my understanding, extension part will be removed from PECAN, PECAN will have seperate Modelling & API17:25
trozetsridhar_ram: tacker seems ok, but neutron looks screwed up17:25
sripriyadiga_: exactly, can you please build the spec along these lines?17:25
sridhar_ramdiga_: yes, can you capture all these points raised .. in the last 15 - 20mins into the spec please ?17:26
diga_sripriya: Sure17:26
diga_sridhar_ram: Sure, will do17:27
sridhar_ramdiga_: great thanks!17:27
diga_sridhar_ram: Welcome!17:27
diga_sridhar_ram: take care! Get well soon!17:27
sridhar_ramsripriya: when u r in the summit and meet infra folks ask them how to create / manage a dev branch for this purpose.. i've no clue17:28
sridhar_ramdiga_: thanks a lot Digambar!17:28
sripriyadiga_: cool, thanks, as sridhar_ram already suggested, you can try fixing smaller RFEs around API layer in parallel as you build the spec, this will help design the spec better.17:28
diga_sridhar_ram: sripriya : can you add your comments in the spec so that I can capture that in next PS17:29
sridhar_ramsripriya: perhaps armax can guide us as well.. they recently created "qos" for classifier work17:29
sripriyasridhar_ram: sure, will note it17:29
sripriyadiga_: will do17:29
*** Vijayendra has quit IRC17:29
diga_sripriya: Yep, I will branch out seperate project in my git repo17:30
diga_sripriya: sridhar_ram: I start my openstack work early morning around 4:30AM everyday because I dont get time in office hours, hope if you will be working late at your tz17:31
sripriyadiga_: i was referring to the main tacker repo itself for API bugs/RFEs, you can also help be a reviewer for API related patches.17:32
diga_sripriya: okay, will review API related paches17:32
sripriyadiga_: thanks! appreciate your commitment and effort!17:34
diga_sripriya: sridhar_ram: yes, will go through the existing API implementation17:35
diga_sridhar_ram: sripriya: welcome! Thanks for giving me the opportunity to work on this task17:35
diga_sripriya: sridhar_ram: its late here, will talk to you on tomorrow17:36
sripriyadiga_: ttyl, gn!17:36
trozetsridhar_ram: can you tell me what your auth_uri is in neutron.conf?17:37
trozetsridhar_ram: oh...openstack network commands work, but not neutron client17:38
*** diga_ has quit IRC17:43
*** yifei has joined #tacker17:51
*** sripriya has quit IRC17:52
*** bobh has quit IRC17:55
*** yifei has quit IRC17:56
*** bobh has joined #tacker17:56
*** sripriya has joined #tacker18:08
*** LamT__ has quit IRC18:11
*** bobh has quit IRC18:18
*** bobh has joined #tacker18:25
*** tung_doan_ has joined #tacker18:31
*** Ravikiran_K has quit IRC18:31
*** vishnoianil has quit IRC18:32
tung_doan_sridhar_ram: seems like network service spec is interesting thing for me... will look into this so that i can support in case of alarm monitoring18:38
sridhar_ramtung_doan_: absolutely.. please review & provide ur comments.. particularly on alarm monitor point of view18:46
tung_doan_sridhar_ram: will do.. thanks :)18:48
sridhar_ramtung_doan_: btw, please continue to help make existing alarm monitor better ... we need func-test for ur feature ASAP18:49
*** tung_doan_ has quit IRC18:53
*** bobmel_ has joined #tacker19:02
*** vishnoianil has joined #tacker19:04
*** bobmel has quit IRC19:06
*** sripriya has quit IRC19:19
*** sripriya has joined #tacker19:37
*** diga has quit IRC19:48
trozetsridhar_ram: ping?19:49
*** Vijayendra has joined #tacker19:55
trozetvishnoianil: ping?19:58
*** achatterjee has quit IRC19:59
*** sripriya_ has joined #tacker19:59
*** sripriya has quit IRC20:00
*** sripriya_ is now known as sripriya20:00
vishnoianiltrozet, heyy20:00
trozetvishnoianil: hi.  I'm hitting an error with  regular networking-sfc: https://paste.fedoraproject.org/443582/raw/20:00
trozetvishnoianil: complaining that hte host_id is none20:00
trozetvishnoianil: seen that before?20:00
*** achatterjee has joined #tacker20:01
vishnoianiltrozet, looking20:02
vishnoianiltrozet, it's while create port-pair-chain?20:02
vishnoianilsorry port-pair20:02
trozetvishnoianil: yeah, what do you normally fill in for hte host_id?20:03
vishnoianiltrozet, did you rebase the driver patch?20:03
vishnoianilbecause host_id is not even part of the API as of now20:03
trozetvishnoianil: i'm using regular networking-sfc OVS driver20:03
trozetvishnoianil: i pulled straight from networking-sfc master20:04
vishnoianilokay looks like it something new that is introduced in sfc api20:04
trozetvishnoianil: i wonder if this is the new host config port binding for neutron20:04
vishnoianilhost_id most-probably be compute_id20:04
*** _achatterjee_ has joined #tacker20:04
vishnoianiltrozet, yeah, but this should not be mandatory20:04
trozetvishnoianil: networking_sfc/services/sfc/drivers/ovs/driver.py:            host_id = port['binding:host_id']20:05
*** achatterjee has quit IRC20:05
trozetvishnoianil: so you never fill out the host_id?20:06
vishnoianilno, my driver is not rebased on the latest networking-sfc code20:07
vishnoianiltrozet, this is related to host interface binding20:08
vishnoianiltrozet, but why is this mandatory field20:08
trozetvishnoianil: yeah the pseduo agent binding20:08
trozetvishnoianil: i think hostconfig type binding is going to be required in ocata20:08
trozetvishnoianil: but not sure why it is in networking-sfc master...if there is no newton branch?20:09
vishnoianiltrozet, i think stephen or cathy can give us more idea about it20:09
vishnoianiltrozet, let me check if there is any existing blurprint for this20:09
trozetvishnoianil: right.  I am sending an email wiht you cc'ed20:09
vishnoianiltrozet, cool, i will rebase my driver today on master branch and push it20:11
trozetvishnoianil: yeah so hte new way you have to configure each OVS with the hardware config20:12
trozetvishnoianil: then ML2 reads it and enteres it into neutron DB20:12
trozetenters*20:12
*** lamt has quit IRC20:12
trozetvishnoianil: crap I just realized i didnt devstack with stable/newton20:14
vishnoianiltrozet, coffee time :D20:15
trozetvishnoianil :)20:15
*** bobh has quit IRC20:27
*** bobh has joined #tacker20:27
*** lamt has joined #tacker20:29
*** sripriya has quit IRC20:36
*** bobmel has joined #tacker20:38
*** bobmel_ has quit IRC20:41
*** bobmel has quit IRC20:55
*** sripriya has joined #tacker21:05
*** tung_doan_ has joined #tacker21:25
*** sripriya has quit IRC21:25
*** tung_doan_ has quit IRC21:29
*** sripriya has joined #tacker21:32
sridhar_ramtrozet: pong21:45
*** sripriya has quit IRC21:55
*** yifei has joined #tacker21:58
*** yifei has quit IRC22:03
*** lamt has quit IRC22:08
*** bobh has quit IRC22:10
vishwanathjlooking for reviewers to review https://review.openstack.org/#/c/380685/22:19
*** sripriya has joined #tacker22:21
*** dkushwaha has quit IRC22:41
*** bobmel has joined #tacker23:07
*** bobmel has quit IRC23:12

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!