Friday, 2016-03-04

« Thursday, 2016-03-03 Index Saturday, 2016-03-05 »
*** gjayavelu has quit IRC00:05
*** pushkaru has quit IRC00:07
*** pushkaru has joined #openstack-glance00:07
*** ayoung has quit IRC00:07
*** pushkaru has quit IRC00:08
*** pushkaru has joined #openstack-glance00:08
*** pushkaru has quit IRC00:11
*** pushkaru has joined #openstack-glance00:14
*** sigmavirus24 is now known as sigmavirus24_awa00:15
*** pushkaru has quit IRC00:27
*** ayoung has joined #openstack-glance01:05
*** achanda has joined #openstack-glance01:14
*** ducttape_ has joined #openstack-glance01:24
*** ducttape_ has quit IRC01:25
*** ducttape_ has joined #openstack-glance01:25
*** ducttape_ has quit IRC01:31
*** mtanino has quit IRC01:37
*** GB21 has quit IRC02:04
*** lin_yang has joined #openstack-glance02:07
*** baojg has joined #openstack-glance02:17
*** ducttape_ has joined #openstack-glance02:25
*** mingdang1 has joined #openstack-glance02:26
*** gjayavelu has joined #openstack-glance02:32
*** gjayavelu has quit IRC02:37
*** haomaiwang has joined #openstack-glance02:39
*** houming has joined #openstack-glance02:44
*** haomaiwang has quit IRC02:45
*** haomaiwang has joined #openstack-glance02:56
*** cdelatte has joined #openstack-glance02:59
*** haomaiwang has quit IRC03:01
*** links has joined #openstack-glance03:01
*** haomaiwang has joined #openstack-glance03:12
*** GB21 has joined #openstack-glance03:13
*** haomaiwang has quit IRC03:17
*** GB21 has quit IRC03:18
*** ducttape_ has quit IRC03:36
nikhiltjcocozz_: +2ed03:39
*** dims has quit IRC03:40
*** ducttape_ has joined #openstack-glance03:42
*** dims has joined #openstack-glance03:43
*** dims has quit IRC03:44
*** links has quit IRC03:44
*** ducttape_ has quit IRC03:56
*** flwang1 has quit IRC03:58
*** gjayavelu has joined #openstack-glance04:13
*** agireud has quit IRC04:27
*** aslaen has joined #openstack-glance04:27
*** links has joined #openstack-glance04:27
*** openstack has joined #openstack-glance04:28
*** haomaiwang has joined #openstack-glance04:28
*** openstack has quit IRC04:30
*** openstack has joined #openstack-glance04:31
*** openstack has quit IRC04:34
*** openstack has joined #openstack-glance04:34
*** agireud has joined #openstack-glance04:34
*** jokke_ has quit IRC04:36
*** openstack has quit IRC04:36
*** openstack has joined #openstack-glance04:36
*** jokke_ has joined #openstack-glance04:37
*** openstack has quit IRC04:39
*** openstack has joined #openstack-glance14:06
*** cdelatte has quit IRC14:08
*** haomaiwang has joined #openstack-glance14:13
*** ducttape_ has joined #openstack-glance14:21
*** ducttape_ has quit IRC14:26
*** khushbu has quit IRC14:28
*** omolchanov1 has quit IRC14:35
*** omolchanov has joined #openstack-glance14:36
*** openstackgerrit_ has quit IRC14:36
*** openstackgerrit_ has joined #openstack-glance14:37
*** smatzek has quit IRC14:39
*** links has quit IRC14:41
jrollflaper87: yep, we're all good now14:42
flaper87mfedosin: could you comment on: https://review.openstack.org/#/c/254334/314:42
flaper87jroll: cool, thanks for the heads up14:42
jrollflaper87: thanks for checking back in :)14:42
openstackgerritTom Cocozzello proposed openstack/glance: no module docs generated  https://review.openstack.org/28808214:44
*** omolchanov has quit IRC14:46
*** omolchanov has joined #openstack-glance14:47
*** ducttape_ has joined #openstack-glance14:56
*** haomaiwang has quit IRC15:01
*** tonyb has quit IRC15:01
*** tonyb has joined #openstack-glance15:01
*** haomaiwang has joined #openstack-glance15:01
*** smatzek has joined #openstack-glance15:03
*** mingdang1 has quit IRC15:05
*** omolchanov has quit IRC15:10
*** omolchanov has joined #openstack-glance15:11
*** mvk has joined #openstack-glance15:22
openstackgerritDarja Shakhray proposed openstack/glance: Adding method "__setitem__"  for class "ImageTarget"  https://review.openstack.org/26689115:25
*** david_lyle__ has joined #openstack-glance15:34
*** david-lyle has quit IRC15:37
*** smatzek_ has joined #openstack-glance15:46
*** mtanino has joined #openstack-glance15:49
*** smatzek has quit IRC15:50
*** groen692 has quit IRC15:52
*** lin_yang has quit IRC15:53
*** dims is now known as dimsum_15:53
*** lin_yang has joined #openstack-glance15:54
*** omolchanov has quit IRC15:58
*** omolchanov1 has joined #openstack-glance15:58
*** vijendar has joined #openstack-glance15:59
*** haomaiwang has quit IRC16:01
*** vijendar1 has joined #openstack-glance16:01
*** vijendar has quit IRC16:04
*** belmoreira has quit IRC16:07
*** khushbu_ has joined #openstack-glance16:10
*** e0ne has quit IRC16:12
*** david_lyle__ is now known as david_lyle16:17
*** pushkaru has joined #openstack-glance16:19
*** achanda has joined #openstack-glance16:19
*** achanda has quit IRC16:20
*** daemontool__ has joined #openstack-glance16:30
*** dshakhray has quit IRC16:33
*** daemontool_ has quit IRC16:34
*** daemontool__ is now known as daemontool16:34
*** lakshmiS has joined #openstack-glance16:42
*** rcernin has quit IRC16:51
*** aslaen has joined #openstack-glance16:53
*** smatzek_ is now known as smatzek16:57
*** ihrachys has quit IRC16:58
*** omolchanov1 has quit IRC17:02
*** omolchanov has joined #openstack-glance17:02
*** mfedosin has quit IRC17:08
*** MattMan has quit IRC17:15
*** MattMan has joined #openstack-glance17:15
*** vijendar1 has quit IRC17:20
openstackgerritMerged openstack/glance: Resolve i18n and Sphinx issues in signature_utils  https://review.openstack.org/28472617:21
*** daemontool_ has joined #openstack-glance17:37
*** vijendar has joined #openstack-glance17:37
*** daemontool has quit IRC17:41
*** vijendar has quit IRC17:44
*** vijendar1 has joined #openstack-glance17:44
*** jistr has quit IRC17:46
*** vijendar1 has quit IRC17:52
*** vijendar has joined #openstack-glance17:55
*** vijendar has quit IRC17:58
*** wxy has quit IRC17:59
*** permalac has quit IRC18:12
*** khushbu_ has quit IRC18:14
*** itisha has joined #openstack-glance18:15
*** khushbu_ has joined #openstack-glance18:18
*** ducttape_ has quit IRC18:26
*** shakamunyi has joined #openstack-glance18:29
*** akanksha_ has quit IRC18:37
*** ducttape_ has joined #openstack-glance18:38
*** pcaruana has quit IRC18:41
*** achanda has joined #openstack-glance18:44
*** vijendar has joined #openstack-glance18:44
*** groen692 has joined #openstack-glance18:51
*** vijendar has quit IRC18:56
*** omolchanov has quit IRC18:56
*** omolchanov has joined #openstack-glance18:56
*** vijendar has joined #openstack-glance18:56
*** smatzek_ has joined #openstack-glance19:02
*** smatzek__ has joined #openstack-glance19:04
*** omolchanov has quit IRC19:05
*** omolchanov1 has joined #openstack-glance19:05
*** smatzek has quit IRC19:05
*** smatzek_ has quit IRC19:07
*** ihrachys has joined #openstack-glance19:10
openstackgerritErno Kuvaja proposed openstack/glance: Sync glance config examples from configgenerator  https://review.openstack.org/28865619:19
*** aslaen has quit IRC19:22
openstackgerritErno Kuvaja proposed openstack/glance: Sync glance config examples from configgenerator  https://review.openstack.org/28865619:22
openstackgerritErno Kuvaja proposed openstack/glance: Update the config files  https://review.openstack.org/28362819:27
flaper87jokke_: re https://review.openstack.org/#/c/288656/ I think bunting had one for this already19:29
* flaper87 needs to find the link19:29
flaper87jokke_: oh, you actually found it19:29
flaper87hahaha19:29
flaper87it's 28865619:29
flaper87rosmaita: around?19:30
rosmaitahey, i'm only 45 sec late19:30
rosmaitahellop19:30
rosmaita*hello19:30
flaper87yoooooooooooo19:31
rosmaitai just put some comments on stuart's import spec patch19:31
rosmaitai haven't had time to write a lot of detail, but we can discuss here where it could go next19:32
rosmaitamy feeling is that we keep the current 3-step-dance proposal, but (maybe) eliminate the extra image states19:32
flaper87ok19:33
rosmaitaoh yeah, for any interested parties, this is what I"m talking about: https://review.openstack.org/#/c/271021/19:33
flaper87mind dropping here your thoughts on ow you thnk we could eliminate it ?19:33
rosmaitasure19:33
rosmaitaso everyone seems to (a) want to use the image resource to track all image-related stuff, and19:34
rosmaita(b) everyone likes the proposal for a "message" on the image resource19:34
rosmaitathe problem with a message is that it's free-form text19:34
rosmaitabut19:34
rosmaitait doens't have to be.19:35
rosmaitai think we could schematize the message field so that it is typed19:35
rosmaitawith each type having a JSON schema describing the content19:35
*** omolchanov1 has quit IRC19:35
*** omolchanov has joined #openstack-glance19:35
rosmaitaso there could be a human readable message, but also easily machine parsed info19:35
rosmaitai wrote up a quick example for the midcycle meeting19:36
rosmaitahttps://etherpad.openstack.org/p/glance-mitaka-midcycle-image-schema-proposal19:36
rosmaitabecause stuart is right, we will need to keep track of hte stuff that a user puts into the "bikeshed"19:37
rosmaitalike checksum, size19:37
*** smatzek__ is now known as smatzek19:37
rosmaitaso we could return those in the "message" field19:37
rosmaita(since we'll be keeping them in the database)19:37
* flaper87 clicks19:37
rosmaitabut the other thing is that the "type" of the message gives us a kind of extra state info19:37
rosmaitawhat i mean is, the user's data that hasn't become an image yet is going to go through a lifecycle19:38
*** ihrachys has quit IRC19:38
rosmaitaand that lifecycle isn't necessarily the same as the image lifecycle19:38
rosmaitabut the user-data lifecycle (uploading, upload complete, processing, done ... something like that)19:39
flaper87rosmaita: what happens when there's a new message ?19:39
flaper87We overwrite the old one19:39
flaper87right?19:39
rosmaitayes19:39
rosmaitabut it's just like queued -> active or something19:39
rosmaitayou dont' know the previous state unless you were looking19:39
flaper87how crazy would it be to have that in a DB table rather than in a free form field ?19:39
flaper87We could have a log for each image19:40
*** pushkaru has quit IRC19:40
flaper87The risk is that it might grow quite a bit19:40
rosmaitaright19:40
rosmaitabut, we could auto-expire19:40
* flaper87 is trying to explore the message idea... not sure how he feels about it yet19:40
rosmaitalike current tasks have an expires_at19:40
rosmaitaso we wouldn't keep an import log around indefinitely19:41
rosmaitayeah, i can't decide if it's genius or a really bad hack19:41
rosmaitabut i am thinking we have a db table for the message (sort of like for tasks), and the schematized message is populated from it19:42
rosmaitabut i was thinking one row per image, not keeping a log19:42
flaper87gotcha19:43
flaper87this would eliminate the need of extra states19:43
flaper87BUT!19:43
* rosmaita waits19:43
flaper87We still need to synchronize states between messages and the image19:43
flaper87I mean, we can't have an uploading image status when the image is killed19:44
flaper87or something like that19:44
flaper87or we probably can19:44
flaper87mmhh19:44
rosmaitano, i think you were right the first time19:44
rosmaitai see it as like the nova task_state thing19:44
rosmaita(which people hate)19:44
rosmaitabut it does give you some visibility into what's going on19:45
rosmaitaso internally, we would have a state of the import that would be checked to see if various image calls should be allowed19:45
rosmaitawe're just not exposing it as a "state", but as a "message_type"19:46
rosmaitaso i do see one drawback19:47
flaper87ok, here's what I'm a bit afraid of about this proposal19:47
* flaper87 waits for the drawback19:47
rosmaitano, you go first19:47
flaper87no you go19:47
rosmaitaok19:47
* flaper87 hopes rosmaita wil lsay "no you go"19:47
flaper87DAMNIT!19:47
rosmaitai'm easy19:47
flaper87dude, haven't you seen Looney toons ?19:48
rosmaitanot in a while19:48
flaper87or whatever that's called19:48
flaper87or Bugs bunny19:48
rosmaitayep, that's looney toons19:48
flaper87"What's up Doc ?"19:48
flaper87ok19:48
flaper87go ahead19:48
flaper87:P19:48
rosmaitathe drawback is that you can't do more than one "external" operation on an image if we use the message field this way19:49
rosmaitafor example, couldn't doa conversion to another image disk_format at the same time you did an export19:49
rosmaitabut since peopel don't like tasks, what can you do?19:49
flaper87lol19:50
rosmaitabut people accept that with nova, you can't do a snapshot and a resize at the same time19:50
rosmaita(pretty sure you can;t)19:50
rosmaitaeven though the server is "active'19:50
flaper87gotcha19:51
flaper87That actually goes along with my current worry19:51
* rosmaita listens19:51
flaper87I'm worried that we will end up creating a limited resource that does something similar to what Stuart recommended but hacked into the image19:51
flaper87I'd hate that to happen.19:51
flaper87not because of Stuart's idea but because hacking it into the image is not ideal19:51
flaper87I'd rather go with Stuart's proposal then19:52
flaper87hope that makes sense19:52
flaper87I'm saying we should be super careful19:52
rosmaitai agree about being super careful19:52
flaper87with what we add to this message thing19:52
flaper87Now, what does this buy us beyond not having to add new states?19:53
flaper87we're fundamentally adding new states anyway19:53
flaper87so, I think it'd be good for us to weight the other benefits19:53
rosmaitawell, the other benefit is that the import-methods are more unified19:53
flaper87I feel like I'm talking like Tarzan. It's just that I'm writing and thinking at the same time. I'm not very good at thinking :P19:54
flaper87or writing19:54
flaper87or anything19:54
flaper87:D19:54
rosmaitame like19:54
rosmaitawell, with stuart's proposal, i'd prefer the one that didn't have free-floating fileIds19:55
flaper87right19:55
flaper87Thing is, I still think the 3-steps import makes sense19:56
rosmaitawell, the problem is, it makes the most sense for the cases we're *not* implementing yet19:56
flaper87right but the goal is to set the basis19:56
flaper87Here's the thing19:56
rosmaitaif you take the direct upload as your paradigm, then import seems ridicuoulsly complicated19:56
rosmaitabut if you think about import form object store, or copy-from, the import call makes a lot of sense19:57
rosmaitaand direct upload seems weird19:57
flaper87I think we should implement direct upload first but that doesn't mean we can't implement swift too in the same cycle. It didn't make sense for Mitaka because we had to design and implement the API19:57
flaper87but we dedicated the whole cycle to designing the API19:57
flaper87which gives enough time to implement most of it in Newton19:57
rosmaitawell, that's a good point19:57
flaper87That said, we should focus on direct upload first19:57
rosmaitai think we have to, for defcore purposes19:58
flaper87right19:58
flaper87but again, I think we can do all this in newton...19:58
rosmaitabut like i said on stuart's patch, i think it will be the least used method19:58
flaper87Newton's PTL will have to be strong at not letting the discussion to start again19:58
flaper87hahahhaa19:58
flaper87rosmaita: it's likely19:59
flaper87rosmaita: have we listed the drawbacks of the 3-steps import process ?20:00
flaper87in the spec, that is20:00
rosmaitanot sure20:00
flaper87if we haven't, I think we should20:00
rosmaitai was so surprised that monty liked it, that i didn't look for drawbacks20:01
rosmaitathat's a good idea20:01
flaper87mordred: ^20:01
flaper87hahahahaha20:01
flaper87mordred: you, sir, are dangerous!20:01
mordredaroo?20:01
flaper87mordred: just talking about the 3-steps import process20:02
rosmaitamordred: you are the father of the glance 3-step image import20:02
mordredit's perfect, right?20:02
mordredwait - it's not implemented yet? :)20:02
rosmaitaif only everyone would see it like you do20:02
flaper87mordred: it shines like a morning after an awful storm that destroyed an entire village20:02
mordredflaper87: dude. those are amazing mornings20:03
flaper87mordred: nope, turns out it took like 6 months of discussion to agree on several things. But newton will be the right cycle for it20:03
mordrednothing like standing in the wreckage of a house destroyed by a tornado with the sun shining down and a cup of coffee in your hand20:03
flaper87mordred: right?20:03
flaper87mordred: wearing your bathrobe20:03
rosmaitayeah, any morning you wake up is a good morning20:03
mordredrosmaita: +10020:03
flaper87mordred: https://specs.openstack.org/openstack/glance-specs/specs/mitaka/approved/image-import/image-import-refactor.html FYI20:03
flaper87it's not urgent. I just thought it'd be fun to let you know you're dangerous... Walking radioactive C420:04
mordredthat's what they call me20:04
mordred"hey, there comes Walking Radioactive C4, let's buy him a beer"20:04
mordredthat's what they all say20:04
rosmaitamordred: actually, if you have time, this is why we're still discussing: https://review.openstack.org/#/c/271021/20:05
flaper87mordred: but there ain't no beer, right?20:05
rosmaitai think you have to buy bourbon for waking radioactive c420:05
flaper87rosmaita: so, what do you think about incorporating drawbacks to the spec. I'll comment on Stuart's review and we can land it20:06
flaper87rosmaita: you should also write another patch to incorporate the message thing20:06
rosmaitaok, i will put up a patch for discussion20:06
rosmaitaok, and also the message thing20:06
flaper87rosmaita: thank you20:06
rosmaitaok, gives my weekend some direction20:07
rosmaitaflaper87: glad we had this discussion, helps to sort things out20:07
*** vijendar has quit IRC20:07
rosmaitaand glad the channel is logged, can stuart an extract20:07
rosmaita*send stuart20:07
tjcocozz_flaper87, have time to talk about https://review.openstack.org/#/c/288082/6/glance/contrib/plugins/image_artifact/v2/image.py ? :-)20:10
flaper87rosmaita: yeah, please! I'm looking forward to hear more from Stuart on this20:10
flaper87tjcocozz_: A-HA! That's your nick!20:11
flaper87tjcocozz_: You are now officially screwed20:11
flaper87:D20:11
flaper87with love from the whole team :D20:11
tjcocozz_flaper87,  oh jeeze that could be a problem :\20:11
flaper87:P20:11
flaper87tjcocozz_: what's up?20:11
tjcocozz_flaper87, so how should i move foward on this patch?  Should i log.E() in the except?20:12
*** lakshmiS_ has joined #openstack-glance20:12
flaper87tjcocozz_: the only thing missing in that patch is. Check if glanceclient is none before calling glanceclient.Client and then raise an error if it is20:13
tjcocozz_flaper87, none of this code is tested :(20:13
flaper87tjcocozz_: no surprise! :( I trust it'll be in Newton with all the rewrite for glare20:14
flaper87tjcocozz_: does my suggestion make sense?20:14
tjcocozz_flaper87, why are we using the image here anyway? seems like a round about way to do something20:14
tjcocozz_flaper87, i probably just don't undrestand since i am new to glance :o20:16
flaper87tjcocozz_: I don't even understand that code... :P20:16
flaper87tjcocozz_: it's the experimental (and old) Artifacts stuff.20:16
flaper87It'll be rewritten in Newton20:16
flaper87which is why I'm not pulling my hair off20:16
*** lakshmiS has quit IRC20:17
flaper87jokke_: https://review.openstack.org/#/c/283628/ <- you need to rebase the checkout of this branch before running configgen20:17
flaper87The change is still pointing to an older head20:17
flaper87jokke_: just `git pull --rebase origin master`20:17
flaper87(from that checkout)20:17
*** vijendar has joined #openstack-glance20:20
*** vijendar has quit IRC20:27
*** pushkaru has joined #openstack-glance20:30
*** khushbu_ has quit IRC20:31
openstackgerritTom Cocozzello proposed openstack/glance: no module docs generated  https://review.openstack.org/28808220:34
tjcocozz_flaper87, thanks for your help!20:36
flaper87tjcocozz_: my pleasure!20:36
flaper87tjcocozz_: and remember, you can't run now20:36
* tjcocozz_ whipes the sweat off his head20:37
tjcocozz_flaper87, btw i am bad at spelling :p   **wipes20:37
flaper87tjcocozz_: what's spelling ?20:39
flaper87:P20:39
*** mfedosin has joined #openstack-glance20:39
tjcocozz_flaper87, luckily you don't need it be that good at it in our jobs \o/20:40
*** vijendar has joined #openstack-glance20:44
*** mfedosin has quit IRC20:48
*** delattec has quit IRC20:58
*** MVenesio has quit IRC21:04
*** smatzek has quit IRC21:21
*** groen692 has quit IRC21:38
*** vijendar has quit IRC21:42
*** dan_nguyen has joined #openstack-glance21:53
*** pcaruana has joined #openstack-glance21:58
dan_nguyenhello, I have a silly question regarding the ability to update the visibility of an uploaded image to True/False for non-admin users.22:00
dan_nguyenIs this supported in Glance?22:01
*** aslaen has joined #openstack-glance22:01
*** vijendar has joined #openstack-glance22:01
openstackgerritMerged openstack/glance: Return 204 rather than 403 when no image data  https://review.openstack.org/25433422:06
*** ninag has quit IRC22:07
dan_nguyennv, looks like glance client is not as current as openstackclient22:07
*** e0ne has joined #openstack-glance22:11
*** aslaen has quit IRC22:17
*** aslaen has joined #openstack-glance22:20
nikhiltjcocozz_: hey22:25
nikhiltjcocozz_: https://review.openstack.org/#/c/288082/7/glance/contrib/plugins/image_artifact/v2/image.py22:25
nikhilthe ImportError will be caught and translated into InvalidArtifactPropertyValue exception22:26
nikhilwhich is not what we should do I think22:27
* nikhil afk22:30
*** pushkaru has quit IRC22:44
*** pushkaru has joined #openstack-glance22:44
*** aslaen has quit IRC22:46
*** vijendar has quit IRC22:55
*** e0ne has quit IRC22:55
*** aslaen has joined #openstack-glance22:57
*** edmondsw has quit IRC23:05
*** e0ne has joined #openstack-glance23:15
*** aslaen has quit IRC23:33
*** itisha has quit IRC23:49
*** ducttape_ has quit IRC23:55
« Thursday, 2016-03-03 Index Saturday, 2016-03-05 »

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