Tuesday, 2013-10-29

« Monday, 2013-10-28 Index Wednesday, 2013-10-30 »
*** yogeshmehra has quit IRC00:13
*** yogeshmehra has joined #openstack-trove00:14
*** yogeshmehra has quit IRC00:18
*** matsuhashi has joined #openstack-trove00:18
*** nosnos has joined #openstack-trove00:34
openstackgerritAuston McReynolds proposed a change to openstack/trove: User-Create Host Does Not Allow Wildcarded Octet  https://review.openstack.org/5421600:57
openstackgerritAuston McReynolds proposed a change to openstack/trove: User-Create Host Does Not Allow Wildcarded Octet  https://review.openstack.org/5421601:02
openstackgerritAuston McReynolds proposed a change to openstack/trove: User-Create Host Does Not Allow Wildcarded Octet  https://review.openstack.org/5421601:09
*** demorris has quit IRC01:18
*** lnxnut has joined #openstack-trove01:20
*** amcrn has quit IRC01:23
*** amytron has quit IRC01:24
*** anteaya has quit IRC01:43
*** haomaiwang has quit IRC01:54
*** haomaiwang has joined #openstack-trove01:54
*** haomaiwa_ has joined #openstack-trove01:58
*** haomaiwang has quit IRC01:59
*** erkules_ has joined #openstack-trove02:07
*** erkules has quit IRC02:09
*** SnowDust has joined #openstack-trove02:52
*** jasonb365 has joined #openstack-trove03:00
*** matsuhashi has quit IRC03:20
*** matsuhashi has joined #openstack-trove03:21
*** SnowDust has quit IRC03:24
*** matsuhashi has quit IRC03:25
*** lnxnut has quit IRC03:43
*** lnxnut has joined #openstack-trove03:43
*** lnxnut has quit IRC03:48
*** cp16net-home has joined #openstack-trove03:52
*** jasonb365 has quit IRC03:55
openstackgerritCraig Vyvial proposed a change to openstack/trove: changing the routes for the api to be clear  https://review.openstack.org/5422404:13
openstackgerritCraig Vyvial proposed a change to openstack/trove: making the bin scripts called with main()  https://review.openstack.org/5422504:13
openstackgerritCraig Vyvial proposed a change to openstack/trove: Fixes trove-api daemon fake mode functionality.  https://review.openstack.org/5126204:13
*** openstackgerrit has quit IRC04:36
*** openstackgerrit has joined #openstack-trove04:36
openstackgerritCraig Vyvial proposed a change to openstack/trove: Fixes trove-api daemon fake mode functionality.  https://review.openstack.org/5126204:41
*** matsuhashi has joined #openstack-trove04:42
*** adrian_otto has joined #openstack-trove04:47
*** yogeshmehra has joined #openstack-trove04:52
*** matsuhashi has quit IRC05:02
*** matsuhashi has joined #openstack-trove05:03
*** demorris has joined #openstack-trove05:06
*** SushilKM has joined #openstack-trove05:06
*** matsuhashi has quit IRC05:08
*** coolsvap has joined #openstack-trove05:09
*** yogeshmehra has quit IRC05:12
*** yogeshmehra has joined #openstack-trove05:12
*** yogeshme_ has joined #openstack-trove05:13
*** yogeshmehra has quit IRC05:13
*** cp16net-home has quit IRC05:28
*** erkules_ is now known as erkules05:39
openstackgerritVipul Sabhaya proposed a change to openstack/trove: Fix bug in Security Group association to Instance  https://review.openstack.org/5423505:41
*** yogeshme_ has quit IRC05:45
*** demorris has quit IRC05:47
*** yogeshme_ has joined #openstack-trove05:48
*** esp has left #openstack-trove06:08
*** SushilKM has quit IRC06:12
*** SushilKM has joined #openstack-trove06:15
*** haomaiwa_ has quit IRC06:18
*** haomaiwang has joined #openstack-trove06:19
*** adrian_otto has quit IRC06:20
*** tanisdl_ has joined #openstack-trove06:26
*** tanisdl has quit IRC06:28
*** yidclare has quit IRC06:28
*** david-lyle has quit IRC06:28
*** tanisdl_ is now known as tanisdl06:28
*** yidclare has joined #openstack-trove06:29
*** david-lyle has joined #openstack-trove06:30
*** SnowDust has joined #openstack-trove06:35
*** haomaiwang has quit IRC06:40
*** haomaiwang has joined #openstack-trove06:40
*** SushilKM has quit IRC06:45
*** yogeshme_ has quit IRC06:52
*** coolsvap has quit IRC07:04
*** denis_makogon has joined #openstack-trove07:05
*** EmilienM has quit IRC07:22
*** denis_makogon_ has joined #openstack-trove07:24
*** ashestakov has joined #openstack-trove07:24
*** EmilienM has joined #openstack-trove07:25
*** denis_makogon has quit IRC07:26
*** coolsvap has joined #openstack-trove07:57
*** ashestakov has quit IRC08:04
openstackgerritIllia Khudoshyn proposed a change to openstack/trove: Initial support for single instance MongoDB support  https://review.openstack.org/5059708:17
*** nosnos_ has joined #openstack-trove08:29
*** haomaiwa_ has joined #openstack-trove08:30
*** SushilKM has joined #openstack-trove08:30
*** nosnos has quit IRC08:30
*** haomaiwang has quit IRC08:32
*** denis_makogon_ has quit IRC08:46
*** nosnos has joined #openstack-trove09:02
*** nosnos_ has quit IRC09:03
*** radez_g0n3 is now known as radez09:57
*** ashestakov has joined #openstack-trove10:05
*** nosnos has quit IRC10:31
*** nosnos has joined #openstack-trove10:32
*** nosnos has quit IRC10:36
openstackgerritnilakhya proposed a change to openstack/trove: Pluggable Extentions using multiple routes  https://review.openstack.org/5428010:38
*** ppenjoy has joined #openstack-trove11:04
*** ppenjoy has left #openstack-trove11:04
*** dukhlov has quit IRC11:06
*** dukhlov has joined #openstack-trove11:06
*** radez is now known as radez_g0n311:31
*** SushilKM has quit IRC11:37
*** SushilKM has joined #openstack-trove11:44
openstackgerritnilakhya proposed a change to openstack/trove: Pluggable Extentions using multiple routes  https://review.openstack.org/5428011:54
*** pdmars has joined #openstack-trove12:03
*** pdmars has quit IRC12:10
*** pdmars has joined #openstack-trove12:20
*** demorris has joined #openstack-trove12:24
openstackgerritDenis M. proposed a change to openstack/trove: Externalisation of heat template  https://review.openstack.org/5431512:32
*** demorris has quit IRC12:33
*** haomaiwa_ has quit IRC12:42
*** haomaiwang has joined #openstack-trove12:42
*** debasish has joined #openstack-trove12:46
*** Barker has joined #openstack-trove13:03
*** jcru has joined #openstack-trove13:08
*** anteaya has joined #openstack-trove13:10
*** SushilKM has quit IRC13:51
*** radez_g0n3 is now known as radez13:52
*** adrian_otto has joined #openstack-trove13:52
*** coolsvap has quit IRC13:54
*** robertmy_ has joined #openstack-trove13:58
*** robertmyers has joined #openstack-trove13:59
*** robertmy_ has quit IRC13:59
*** robertmy_ has joined #openstack-trove14:02
*** robertmyers has quit IRC14:02
*** robertmy_ is now known as robertmyers14:03
*** grapex has joined #openstack-trove14:03
*** amytron has joined #openstack-trove14:04
*** lnxnut has joined #openstack-trove14:30
*** amytron has quit IRC14:34
*** amytron has joined #openstack-trove14:35
*** datsun180b has joined #openstack-trove14:37
openstackgerritCraig Vyvial proposed a change to openstack/python-troveclient: adds support for configurations management  https://review.openstack.org/5316914:49
ikhudoshyncp16net: ping14:50
cp16netikhudoshyn: hi14:50
cp16netwhoops14:50
cp16netpong14:50
ikhudoshyncp16net: pls find my answer on ur comment https://review.openstack.org/#/c/53378/14:51
cp16netahhhh14:51
cp16netuntil conductor changes that..14:51
cp16net:-P14:51
*** jasonb365 has joined #openstack-trove14:53
cp16netdmakogon: ping14:54
ikhudoshyni look forward to when it happens too))14:54
cp16netyeah it looks close to being done14:54
cp16netrather merged14:55
dmakogoncp16net, pong14:58
dmakogoncp16net, 'sup ?14:58
*** coolsvap has joined #openstack-trove14:58
*** openstackgerrit has quit IRC15:00
*** openstackgerrit has joined #openstack-trove15:01
*** SnowDust has quit IRC15:07
*** adrian_otto has quit IRC15:07
openstackgerritIllia Khudoshyn proposed a change to openstack/trove: Fix issue with PBR  https://review.openstack.org/5435215:10
*** jasonb365 has quit IRC15:12
dmakogonrobertmyers, ping15:16
robertmyersdmakogon: pong15:16
dmakogonhttps://review.openstack.org/#/c/54315/15:16
dmakogonrobertmyers, and also i answered you ML thread15:16
*** jasonb365 has joined #openstack-trove15:17
robertmyerscool, the review looks nice15:18
dmakogonrobertmyers, have you received email ?15:20
robertmyersI haven't even gotten mine yet :)15:20
robertmyersemail is ta slow15:20
robertmyers:)15:20
dmakogonrobertmyers, i need to hear your thoughts about placement strategy15:22
cp16netdmakogon: hmm so casandra needs the jdk huh?15:23
dmakogoncp16net, yes15:24
cp16netthat sounds odd15:24
dmakogoncp16net, and usage of jdk with cassandra is production way15:24
robertmyersdmakogon: I would like it if there was just one template directory. But I could be persuaded to have them be separate15:24
cp16neti've not used casandra but when i was reading over it seems odd15:24
cp16netbut ok15:24
cp16neti'll remove the -115:25
dmakogoncp16net, thanks15:25
robertmyersthe jinja get_tempates can look in mulitple directories too15:25
dmakogonrobertmyers, using suggested by you placement strategy is not so flexible15:26
dmakogonrobertmyers, because you need to modify EVN every time you bring new datastore support15:26
dmakogonrobertmyers, suggested by me is a bit global15:27
robertmyersno, you can do ENV.get_template('{data_store}/foo')15:27
robertmyersor ENV.get_template('config/foo')15:28
robertmyersor ENV.get_template('heat/foo')15:28
dmakogonrobertmyers, {value} - could be used ?15:28
robertmyersno, you'd have to do string replacement15:29
dmakogonrobertmyers, jinja could work with dynamic variables ?15:29
robertmyerslike this?15:29
robertmyersor ENV.get_template('%s/foo' % data_store)15:29
dmakogonrobertmyers, i thought that jinja could work with routs without string replacements15:30
robertmyersI'll have to consult the docs a bit15:30
dmakogonrobertmyers, i've spent some time on reading jinja sources, and i understood that jinja could work with complete path, it can't work with dynamic routes15:32
*** adrian_otto has joined #openstack-trove15:33
robertmyersbut you could use select_template15:33
robertmyershttp://jinja.pocoo.org/docs/api/#jinja2.Environment.select_template15:33
robertmyersand pass a number of template paths15:33
*** tanisdl has quit IRC15:34
*** radez is now known as radez_g0n315:34
*** paul_lodronio has joined #openstack-trove15:34
dmakogonrobertmyers, i suppose such select is the same as get15:34
dmakogonrobertmyers, we have no validation for templates15:34
robertmyersENV.select_template(['data_store/config', 'data_store.config'])15:35
*** radez_g0n3 is now known as radez15:35
robertmyersis what I'm thinking15:35
dmakogonrobertmyers, it would bring us to duplicating15:36
robertmyersthen you could have a flat structure or nested15:36
robertmyersno, if finds the first one15:36
robertmyersso you go in order of most specific15:36
dmakogonrobertmyers, as, i suppose, you've seen i'm using paths template defined through config15:36
*** SushilKM has joined #openstack-trove15:37
dmakogonrobertmyers, you are trying to protect yourself from missing template, but idea is to use more global placement strategy that would not be harmful for Trove15:38
dmakogonrobertmyers, missing configuration template is not so critical, but missing heat template would lead us to big epic fail15:39
robertmyerswhy would it be harmful to just use the directory name in the get_template?15:39
dmakogonrobertmyers, we have only one path for templates15:39
robertmyersdmakogon: yes so inside of that place configuration and heat15:40
robertmyersthen in get_template('heat/data_store.heat.template')15:41
dmakogonrobertmyers, but i don't want to use path builder through %15:41
robertmyersi'm confused because you already are?15:41
robertmyerstemplate_filename = "%s.heat.template" % service_type15:42
robertmyersinstead template_filename = "heat/%s.heat.template" % service_type15:42
dmakogonanyway, i'd like to have it like i did15:43
robertmyersdmakogon: i'd like to have it my way too :)15:44
robertmyersbut this is a really good start15:45
*** haomaiwang has quit IRC15:45
robertmyersI'd be okay with it15:45
dmakogonrobertmyers, thanks15:45
*** haomaiwang has joined #openstack-trove15:46
*** aignatov has quit IRC15:53
openstackgerritAndrey Shestakov proposed a change to openstack/trove: Add support of datastore types  https://review.openstack.org/4793415:54
dmakogonrobertmyers, ping again)15:54
dmakogonrobertmyers, about your nit))) point is that we could store template anywhere we want. That is what is thought when i was writing this code.15:55
*** jrodom has joined #openstack-trove16:02
*** robertmy_ has joined #openstack-trove16:06
*** robertmyers has quit IRC16:06
*** SushilKM has quit IRC16:06
robertmy_dmakogon: I'm just suggesting that we have one custom path for all templates16:08
*** robertmy_ is now known as robertmyers16:08
robertmyersyou could still put that anywhere you wanted16:09
dmakogonrobertmyers, i got it, but would be a bit flexible to let admin decide where to store templates ?16:09
dmakogonrobertmyers, in your case i should store whole directory together16:09
robertmyersdmakogon: no we could have sub directories to store the various templates16:10
robertmyerseither by type or by data_store16:10
robertmyers /etc/trove/templates/heat16:10
robertmyers /etc/trove/templates/configuration16:10
robertmyersor16:10
robertmyers /etc/trove/templates/data_store/16:11
robertmyers /etc/trove/templates/mysql/config16:11
robertmyers /etc/trove/templates/mysql/heat16:11
ikhudoshyngrapex: hi, could u re-approve this pls https://review.openstack.org/#/c/53378/16:14
grapexikhudoshyn: Done16:16
ikhudoshyngrapex: would be great if u re-approve this too (at least when robots +1 it) https://review.openstack.org/#/c/50597/ -- I rebased it against the latest master16:16
ikhudoshyngrapex: tnx16:17
grapexvipul SlickNik: Any chance you guys could look at some of these? https://review.openstack.org/#/c/45116/ , https://review.openstack.org/#/c/53165/ , https://review.openstack.org/#/c/53381/16:17
ikhudoshyngrapex: btw, do we know what to do with broken integration?16:17
grapexikhudoshyn: I don't.16:17
*** jasonb365 has quit IRC16:27
openstackgerritA change was merged to openstack/trove: Add Backup/Restore Checksum Validation  https://review.openstack.org/5316516:35
*** jasonb365 has joined #openstack-trove16:38
*** jasonb365 has quit IRC16:40
juicegrapex: you appear to be the sole approver today...this is a minor change but a major fix that is holding up a deployment ->https://review.openstack.org/#/c/5423516:55
juicesecurity groups are not getting assigned to instances - this is a regression bug16:56
*** esp has joined #openstack-trove16:56
vipulanyone know why the gate is broken? seem to be failing to upload image16:57
vipulhttps://gist.github.com/vipulsabhaya/721853116:57
vipulI recall some disucssion on this yesterday16:57
vipulwas there a fix?16:58
robertmyersvipul: not yet16:58
vipulrobertmyers: thx17:00
vipulgrapex: you'll have to 'trust me' that this works :) https://review.openstack.org/#/c/5423517:00
vipulgrapex: need it approved today if possible17:00
openstackgerritDenis M. proposed a change to openstack/trove: Absence of conf template issue  https://review.openstack.org/5387617:01
grapexvipul: Oh no, I believe you. :)17:03
grapexvipul: I +2'd, go ahead and approve.17:03
vipulawesome thanks grapex17:03
datsun180bi hear core is bestowing boons today17:04
grapexvipul: no problem17:04
datsun180bwhere do i get in line?17:04
grapexvipul SlickNik: Now, if one of you could look at Conductor: https://review.openstack.org/#/c/45116/17:05
openstackgerritDenis M. proposed a change to openstack/trove: Datastore backup flexibility  https://review.openstack.org/5290517:05
vipulgrapex: Ok i owe you one, so that's next on my list :)17:05
dmakogonguys, could anyone review this one ? https://review.openstack.org/#/c/52666/17:07
dmakogonand this one https://review.openstack.org/#/c/51884/17:07
openstackgerritA change was merged to openstack/trove: Fix bug in Security Group association to Instance  https://review.openstack.org/5423517:08
openstackgerritDenis M. proposed a change to openstack/trove: Externalisation of heat template  https://review.openstack.org/5431517:09
ikhudoshynvipul: ping17:10
*** ashestakov has quit IRC17:10
ikhudoshynvipul: https://review.openstack.org/#/c/53378/ , pls, that was ur -2, now I hope I made it better17:11
openstackgerritDenis M. proposed a change to openstack/trove: Absence of conf template issue  https://review.openstack.org/5387617:11
*** haomaiwang has quit IRC17:12
*** haomaiwang has joined #openstack-trove17:13
dmakogonvipul, could you please re-approve this one (i rebase it)? https://review.openstack.org/#/c/52905/17:15
juicethanks grapex!17:16
*** yogeshmehra has joined #openstack-trove17:22
dmakogonGuys, i suppose we are ok with one heat template per datastore type for single instance ?17:38
yogeshmehradmakogon: do we need to discuss on AWS vs OS expressions?17:40
dmakogonyogeshmehra, i research a bit, there is no OS resource for instance and volume17:41
dmakogonyogeshmehra, so, not there is nothing to discuss17:41
yogeshmehraok...17:42
dmakogonhttps://github.com/openstack/heat/blob/master/heat/engine/resources/instance.py#L611-L61217:42
dmakogonhttps://github.com/openstack/heat/blob/master/heat/engine/resources/volume.py#L428-L42917:42
dmakogonoh, we need to update volume resource17:42
dmakogonyogeshmehra, example https://github.com/openstack/heat-templates/blob/f30713405b1221d86bb6c4af3810597a32a5aca1/hot/F18/NovaInstanceWithCinderVolume_Native.yaml#L48-L5817:44
dmakogonyogeshmehra, are you working on heat template ?17:45
*** amcrn has joined #openstack-trove17:46
dmakogonyogeshmehra, ping17:47
SlickNikesp: rdjenkins gate is broken because of this https://bugs.launchpad.net/ubuntu/+source/python-pbr/+bug/124567617:55
SlickNikfyi17:55
*** ashestakov has joined #openstack-trove17:55
SlickNikOnce a fix for that gets pushed out, things should start working.17:55
espthx SlickNik17:55
*** jmontemayor has joined #openstack-trove17:55
espthought it was me :)17:55
*** Barker has quit IRC17:56
dmakogonlol17:58
dmakogonany idea how to avoid this ?17:58
dmakogoneveryone should use 0.5.2117:58
dmakogonor someone who did bad commit should be punished after he fixes it18:00
*** Barker has joined #openstack-trove18:01
dmakogonhm, i have same ubuntu, but i didn't got error18:01
dmakogonstrange18:01
*** SushilKM has joined #openstack-trove18:05
yogeshmehradmakogon: had to step into a meeting18:08
yogeshmehrasushilKM is working on heat template18:09
amcrnSlickNik: thanks for the info18:12
*** Barker has quit IRC18:12
*** SnowDust has joined #openstack-trove18:22
*** jasonb365 has joined #openstack-trove18:24
*** jasonb365 has quit IRC18:25
*** jasonb365 has joined #openstack-trove18:27
*** amytron_ has joined #openstack-trove18:32
*** SnowDust has quit IRC18:32
*** Barker has joined #openstack-trove18:32
*** amytron has quit IRC18:33
*** amytron_ is now known as amytron18:33
*** SnowDust has joined #openstack-trove18:45
*** Barker has quit IRC18:46
*** Barker has joined #openstack-trove18:51
*** jasonb365 has quit IRC18:56
*** jasonb365 has joined #openstack-trove18:58
*** jasonb365 has quit IRC19:00
*** jasonb365 has joined #openstack-trove19:01
*** anteaya has quit IRC19:03
*** anteaya has joined #openstack-trove19:08
*** denis_makogon has joined #openstack-trove19:12
*** SushilKM has quit IRC19:14
*** SushilKM has joined #openstack-trove19:15
*** coolsvap has quit IRC19:17
denis_makogonanybody home ?19:33
datsun180brax is in meetings19:34
yogeshmehradmakogon: there?19:38
*** SushilKM has quit IRC19:47
*** vipul is now known as vipul-away19:49
denis_makogonyogeshmehra, yes19:58
yogeshmehradenis_makogon: https://review.openstack.org/#/c/54315/19:58
yogeshmehrai did not get the intent19:58
yogeshmehrais it not duplicate19:58
denis_makogonyes this one is not a dublicate19:59
denis_makogonthis one is implemented in another way suggested by community19:59
yogeshmehraok, but community has been giving comments...on the patchset20:00
yogeshmehrawhy a new patchset?20:01
SnowDustwhy a separate commit :-)  ?20:04
denis_makogonyogeshmehra, because i'd like to community to decide which one is better, and SnowDust was taking to long to answer all comments20:04
denis_makogonSnowDust, you still able to challange me commit20:04
denis_makogon*my20:04
*** harlowja has quit IRC20:04
yogeshmehradenis_makogon: the challenge is, why a new commit? :-)20:05
denis_makogonyogeshmehra, SnowDust have you recieved ML with robertmyers complains ?20:05
*** harlowja has joined #openstack-trove20:05
*** vipul-away is now known as vipul20:05
denis_makogonyogeshmehra, stackanalytics.org20:05
SnowDustdenis_makogon : jinja templating .. point was taken !20:05
denis_makogongood20:06
SnowDustand understood .. with ur comments on it20:06
SnowDustwhen u disagreed .. citing community .. and then accepting .. later20:06
*** mmcdaris has joined #openstack-trove20:06
denis_makogonSnowDust, preasure20:06
SnowDusti was following it .. throughout ..20:06
yogeshmehraSnowDust: i think the point which is valid is, the delay in responses...20:06
denis_makogonpersonaly, your patchset is better20:07
denis_makogonjinja is heavy tank against this task20:07
SnowDustyeah i have a brilliant co-author .. Denis M. working with me :-) has to be !20:07
denis_makogonand jinja doing a lot of tasks we really don't nead20:07
denis_makogon*need20:08
SnowDust:-) then why did u code it ?20:08
denis_makogonSnowDust, true, i would add you to Co-author20:08
denis_makogonSnowDust, to see reaction20:08
SnowDustno .. need !20:08
SnowDustlets conserve energy20:08
denis_makogonSnowDust, we still could fight for you patchset20:08
denis_makogonML !!20:08
denis_makogonnot in chat20:09
denis_makogonML20:09
denis_makogonand only20:09
SnowDustunderstand ..20:09
denis_makogonrobertmyers already started topic for this question20:10
denis_makogonSnowDust, have you subscribed on openstack-dev ?20:11
vipuldenis_makogon: competing patches are not cool - please work with the original author to rework a patchset, not submit a new one cuz maybe you thikn your's is better20:11
denis_makogonvipul, i know that, the reason why i didn't push new patchset into SnowDust review is that they are a bit different20:12
vipuldenis_makogon: what happens if merge both patchsets?20:12
denis_makogonvipul, we would have another one way to load templates20:13
denis_makogonpersonaly i don't want to use jinja for this task20:13
vipuldenis_makogon: they implement the same BP - and take two different approaches20:14
vipuldenis_makogon: sure, then make a case for why not, and let the original author change it20:14
denis_makogonvipul, because we need it fast, this task has status "emergencу"20:17
denis_makogonvipul, i suppose we need to range estimate for iit20:17
denis_makogonSnowDust, you could take my whole patchset and push it to your review20:18
denis_makogonSnowDust, estimate is 1d maximum20:18
SnowDustok i will finish my side of it in few hours from now20:19
yogeshmehradenis_makogon: ur patchset, was it worked uponthe review comments on the existing patchset?20:20
denis_makogonyes20:20
denis_makogon100%20:20
yogeshmehrabut then you are syaing u dont agree with that20:21
denis_makogonyogeshmehra, me - i don't like it20:21
denis_makogoncommunity what's it that way20:21
*** radez is now known as radez_g0n320:22
yogeshmehrasure...20:22
denis_makogonhuge problem is that noone fights for his own review20:22
SnowDusti did say .. that i am limited by the scope of BP ..20:23
SnowDustalso did ask the PTL to orient the reviewers ..20:24
denis_makogonso, in this case could win the race worth review from others ...20:24
*** jasonb365_ has joined #openstack-trove20:24
denis_makogonjinja is bad dependency20:24
SnowDustthats why i opposed it when we were discussing the BP20:24
robertmyersdenis_makogon: I disagree20:24
robertmyersfor loading it is much better20:25
denis_makogonrobertmyers, in proccessor tick jinja looses20:25
robertmyersi don't wish to replace the logic of the templates20:25
SnowDustjinja is not dependancy here .. but twist :-)20:25
robertmyersjinja is just as fase as reading from the os20:25
*** jasonb365 has quit IRC20:25
*** jasonb365_ is now known as jasonb36520:25
robertmyersjinja caches the templates20:25
SnowDustwe dont have to twist it .. if it can be done with simplest code pieces which themselves are not complicated20:25
denis_makogonrobertmyers, not only reading20:26
denis_makogonrobertmyers, we do we need to cache templates ?20:26
SnowDusthow caching helps here ?20:26
SnowDustwhen . .the instance creation time is 1000x times more than the template fetch activity ?20:26
robertmyersso you don't have to read from the file system each template read20:26
openstackgerritSteve Leon proposed a change to openstack/trove: Adding designate dns support to trove  https://review.openstack.org/5441220:27
denis_makogoncaching is advantage only at HighLoad systems20:27
robertmyersspeed isn't the issue is all i'm saying20:27
SnowDustrobertmyers : why will we cache it here .. when . .the instance creation time is 1000x times more than the template fetch activity20:27
robertmyersJinja will not slow it down20:27
denis_makogonrobertmyers, same as os20:28
denis_makogonwhat if someday jinja would brake ?20:28
robertmyers>> robertmyers, in proccessor tick jinja looses20:28
vipulthe advantage is that you have a consistent code path for loading dependencies like this20:29
vipulwe should consider even doing this for the user_data we load20:29
robertmyersdenis_makogon: what if our simple logic to read from the os has bugs20:29
robertmyersvipul: +120:30
denis_makogonvipul, you not convinced me20:30
vipulOk simple example.. say you want to include the heat template in Trove package, instaed of on the file system20:31
vipulhow does the current proposal handle that..20:32
vipulit seems that we have a different choices for where to put say a my.cnf template.. but if you want to load a heat template, it better be on the file system20:32
robertmyersvipul: that is my point20:33
denis_makogonvipul, we have mandatory  path, package is also directory in terms of operating system20:33
robertmyerswe gain the sane loading tools and lose nothing20:33
denis_makogonin this case i would like to see optinal algorithms for config loading20:33
*** mayfield has joined #openstack-trove20:35
vipuli dont' like have different code paths, and having to know that you can only load templates20:36
vipulif it's on the filesystem20:36
*** radez_g0n3 is now known as radez20:36
denis_makogoni quit this discussion until SnowDust will update his review20:37
denis_makogonif i made this review i should want to get it merge, ok, concept of my bp is perfect, jinja is our everything, ok))20:39
denis_makogonwaiting for comments on review20:39
denis_makogonand also it would be nice to get my reviews reviewed20:40
vipulthere is value in simplifying our code base so deployers don't need to keep track of every minute detail20:43
vipulwhat robertmyers is suggests goes a long way towards that20:43
denis_makogonk20:44
yogeshmehraso what is gather is: use jinja to load the template...(so that it can be included in the package as other conf files do)...AND the have the template parameters handle the service-type hardcoding20:44
yogeshmehrawhat i gather*20:45
vipulyogeshmehra: Yes, we don't have to use Jinja to do what Heat does.. as what's been discussed in the email thread20:45
vipulbut let's go with a common mechanism for loading these types of dependencies20:45
yogeshmehrathis will enable us to be consistent...and then load the same template for any number of services (depending on how geneirc we can make our templates))20:46
SnowDustyogeshmehra: update the BP .. so reviewers have clarity20:46
SnowDustwith this new discussion20:46
yogeshmehrarobertmyers: does this suffice?20:47
denis_makogonyogeshmehra, do not parametrise template with service type, PLEASE !!!20:47
denis_makogon1 template per service type20:47
robertmyersyogeshmehra: yes20:48
yogeshmehraso this is the point of contention still...20:48
denis_makogonyogeshmehra, i suppose we discuss it 3 or more ti,es20:48
yogeshmehrayes we did./..20:48
denis_makogon*times20:48
yogeshmehrai agree...20:48
denis_makogonstatic templates, fixed parameters, no modifying20:48
yogeshmehradenis_makogon: i know what u r saying20:49
SnowDustservice type will be there in template itself .. and template has its own unique name20:49
yogeshmehrai think we discussed this in detail;20:49
robertmyersdenis_makogon: yes, that is what I want too20:49
denis_makogonrobertmyers, <320:49
robertmyersyou can even wrap it in {% raw%} {% endraw %}20:49
robertmyersto make it so20:50
yogeshmehramysql percona, will have same template...20:50
yogeshmehrawould they be just the replica..20:50
yogeshmehraexcept for hardcoded servicetype20:50
yogeshmehrai mean same template content but different templates, is it?20:51
denis_makogonyogeshmehra, soon percona would desapear as service type20:51
SnowDustthats why i hated templating .. as admin perspective ( even if i did a lot of django ) .. they have to understand the noodles and chopsticks20:51
yogeshmehradenis_makogon: ok...20:51
yogeshmehraSnowDust: lets stay on topic... :-)20:51
denis_makogonSnowDust, yes, that is their job20:51
datsun180bi don't know that i understand the noodles and chopsticks of the phrase "understand the noodles and chopsticks"20:52
yogeshmehra"nuts and bolts" :-)20:52
denis_makogondatsun180b, they should know heat20:52
SnowDustyogeshmehra: replied to robertmyers on {% raw%} {% endraw %}20:52
yogeshmehraoh...silly me20:52
yogeshmehra:-)20:52
yogeshmehraso is the consensus only on....using the jinja for loading...20:53
yogeshmehra?20:53
yogeshmehraplz bear with me... :-)20:53
denis_makogonyogeshmehra, what are your estimates to finish ?20:53
yogeshmehraSnowDust wanted to finish it in next hour or so20:53
denis_makogonyogeshmehra, only jinja20:53
denis_makogonyogeshmehra, i hope20:54
yogeshmehrabut given the timezone he is in...20:54
yogeshmehramay be tomorrow20:54
denis_makogonk20:55
denis_makogonbut we need it fast, i suppose20:55
yogeshmehradenis_makogon: completely agree +120:55
denis_makogonSnowDust, when you finish ping me and i'll close my review20:56
grapexSnowDust: "Noodles and chopsticks," I like that phrase, I think I'm going to start stealing it. :)20:57
yogeshmehraby the way, on AWS vs OS, i think IS is still incomplete, per denis_makogon's point earlier20:57
SnowDustsure !20:57
yogeshmehraIS -> OS20:57
SnowDustgrapex : meant .. when u are hungry .. u dont need to learn the food and the new tool to relish it !20:57
denis_makogonyogeshmehra, i thought this topic closed20:57
yogeshmehradenis_makogon: just double checking20:57
denis_makogonyogeshmehra, only volume has OS resource20:57
SnowDustsame is templating with heat concept :D20:58
grapexSnowDust: Ah! I like that even better then. :)20:58
* denis_makogon gone to watch new Elementary episode20:59
*** demorris has joined #openstack-trove20:59
SnowDustgrapex: dont use in summit ! .. u may be deported hehe20:59
grapexSnowDust: The US government has already bribed China to ensure they can get rid of me for at least a week.21:00
grapexI think the next Summit should take place on international waters at a semi-disclosed location in the middle of the Pacific.21:01
*** sbisbee_ has joined #openstack-trove21:01
SnowDustgrapex: Guantanamo ?21:01
grapexVarious companies will send cargo or cruise ships and circle them.21:01
grapexSnowDust: That could work too! I heard the beaches there are beautiful.21:01
grapexhub_cap actually lived there at one point in his life, you should ask him about it.21:02
yogeshmehradenis_makogon: isn't OS:Nova:Server sufficing instance expression in AS?21:02
SnowDusthttp://docs.openstack.org/developer/heat/template_guide/openstack.html#OS::Nova::Server21:02
yogeshmehraAWS?*21:02
*** saurabhs has joined #openstack-trove21:02
SnowDustthis is equivalent of http://docs.openstack.org/developer/heat/template_guide/cfn.html#AWS::EC2::Instance21:02
SnowDustbut not 100% compatible ..21:02
denis_makogonyogeshmehra, did you look at heat code ?21:02
SnowDustand thats why modred / robertmyers suggested a change21:02
*** pdmars has quit IRC21:03
vipulwhat is missing in OS templates that is implemented in AWS21:03
yogeshmehrawell...i did bit of dry run around the provisioning21:03
yogeshmehrabut whats the point?21:03
yogeshmehravipul: yeah, we ar eon that topic...21:03
robertmyersSnowDust: I actually don't care about the actual template contents, just was wondering why not21:04
SnowDustAWS::CloudFormation::Init < missing in OS::nova::server21:04
denis_makogonSnowDust, +121:04
SnowDustbut can be overcome using UserData21:04
yogeshmehrayeah...21:04
vipulok.. so there is a workaround21:04
SnowDustyeah ..21:05
SnowDustneed to test that ..21:05
yogeshmehraSnowDust: we need tot est that21:05
yogeshmehratest that*21:05
yogeshmehracrazy finger syndrome...sorry21:05
denis_makogonwe need heat-master21:05
yogeshmehra:-)21:05
vipulSo if it will actually work, then why not implement it correctly the first time?21:06
SnowDustvipul: right .. trying that21:06
vipulSnowDust: Ok - just checking.. i heard opposition so just clarifying21:06
SnowDustvipul: i did oppose coz it was around the template .. rather around the refactoring i did :-)21:07
*** erkules has quit IRC21:07
*** westmaas has quit IRC21:07
*** sbisbee has quit IRC21:07
*** hub_cap has quit IRC21:07
SnowDustbut .. its for the benefit of community .. so going as per community :-)21:07
yogeshmehradenis_makogon: heat-master?21:09
vipulSnowDust: +121:10
denis_makogonyogeshmehra, a guy who knows everything about heat21:10
*** amytron has quit IRC21:11
yogeshmehradenis_makogon: funny but i thought its a component i don't know  about, which gives extra-terrestrial apis for heat...21:12
yogeshmehra:-)21:12
*** amytron has joined #openstack-trove21:12
*** erkules has joined #openstack-trove21:14
*** westmaas has joined #openstack-trove21:14
*** mmcdaris has quit IRC21:14
*** hub_cap has joined #openstack-trove21:16
*** lnxnut has quit IRC21:19
*** radez is now known as radez_g0n321:22
*** mmcdaris has joined #openstack-trove21:24
*** paul_lodronio has left #openstack-trove21:26
esmuteAny ideas what is wrong with rd-jenkins?21:32
openstackgerritJenkins proposed a change to openstack/trove: Updated from global requirements  https://review.openstack.org/5443321:33
amcrnesmute: [10:55:08] <SlickNik> esp: rdjenkins gate is broken because of this https://bugs.launchpad.net/ubuntu/+source/python-pbr/+bug/124567621:34
esmuteThanks amcrn21:35
datsun180boh don't even joke about that update Jenkins21:35
amcrnnp21:35
*** jasonb365 has quit IRC21:35
espthx amcrn !21:35
*** harlowja has quit IRC21:36
*** harlowja has joined #openstack-trove21:37
*** mayfield has quit IRC21:38
grapexdatsun180b: https://review.openstack.org/#/c/54436/21:50
clarkbgrapex: about that, how are they working now?21:50
grapexclarkb: The Trove tests?21:51
datsun180bgrapex: yeah i saw that21:51
clarkbgrapex: ya21:51
datsun180bi figure i can't hurt your review with my +121:52
grapexclarkb: Well enough- there is different problem affecting the "reddwarf" gate which runs against a fully stood up VM, but the tests which run with Tox seem to have no problems.21:53
datsun180bthat problem has something to do with pbr 0.5.22 crying during our disk-image-builder process iirc21:54
*** yogeshmehra has quit IRC21:54
clarkbdatsun180b: is that problem related to 54436? I am trying to understand why that change is necessary when presumably the tests have been running previously. Were we just not running the tests in openstack infra land previously?21:55
datsun180bclarkb: sorry, unrelated21:56
grapexclarkb: Are the Tox classified as part of "Infra Land?"21:56
clarkbgrapex: sure as we run tox in infra land21:56
grapexclarkb: Ok, then yes. Here's an example: https://review.openstack.org/#/c/45116/21:56
datsun180b54436 is something we need, i noticed at the end of friday that git-review seemed to have updated our requirements for my patchset, assumedly from global-reqs21:57
datsun180bamong those changes somehow probo was removed from my reqs21:57
clarkboohhhh it was grandfathered in, I get it now21:57
datsun180bhttps://review.openstack.org/#/c/45116/16..17/test-requirements.txt21:58
datsun180bthing is, i didn't ask for those reqs to be updated right then, i have no idea what triggered it21:58
clarkbdatsun180b: rebase probably21:59
clarkbgit log should be able to confirm22:00
grapexclarkb: So, without that we can't proceed. Is there a problem with merging it?22:00
clarkbgrapex: nope, I just wanted to make sure I understood why it suddenly needed to go in. I get it now22:00
grapexclarkb: Yeah, sorry I missed it earlier.22:00
grapexdatsun180b: And sorry I didn't really get what you were trying to tell me when you brought this up yesterday. :(22:01
datsun180bit's okay, 45116 ps 19 has the reqs it needs22:01
grapexdatsun180b: I just thought it was another one of your tall tales!22:01
datsun180bi'm learning that you have some kind of allergy to in medias res22:02
*** Barker has quit IRC22:02
datsun180bat least when it comes to my stories22:03
grapexdatsun180b: I like it for films and plays, just not when you're explaining a technical issue. :)22:04
datsun180bi'll take care to cover my context better in the future22:05
*** robertmyers has quit IRC22:07
*** harlowja has quit IRC22:08
*** jodom has joined #openstack-trove22:19
*** SnowDust has quit IRC22:20
*** jrodom has quit IRC22:21
*** denis_makogon has quit IRC22:21
*** mmcdaris has quit IRC22:30
grapexvipul SlickNik: Any update on this? https://review.openstack.org/#/c/45116/22:45
*** datsun180b has quit IRC22:48
vipulgrapex: sorry have not had a chance, promise to look though22:51
vipultoo much firefighting22:51
*** jcru has quit IRC22:52
grapexvipul: Well, I can sympathize with you on that one. Just try to look at it soon-ish, it would be nice if Datsun180b could get out from under it for awhile.22:53
*** harlowja has joined #openstack-trove23:08
*** grapex has quit IRC23:17
openstackgerritDan Nguyen proposed a change to openstack/trove: Add optional ip address filter for hiding ips  https://review.openstack.org/5376623:26
*** jmontemayor has quit IRC23:36
*** yogeshmehra has joined #openstack-trove23:36
*** amytron has quit IRC23:40
*** demorris_ has joined #openstack-trove23:49
*** jcooley_ has joined #openstack-trove23:57
*** demorris has quit IRC23:58
*** jesusaurus has quit IRC23:58
*** jcooley has quit IRC23:58
*** demorris_ is now known as demorris23:58
*** jcooley_ is now known as jcooley23:58
« Monday, 2013-10-28 Index Wednesday, 2013-10-30 »

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