Thursday, 2017-02-02

rosmaitahemanthm: about that change you made to eliminate the constants ... i can't find get_alembic_branch_head in glance.db.sqlalchemy00:04
hemanthmrosmaita: it's here
rosmaitaty, i  was looking at an old version00:05
hemanthmI just realized that the number of rows we return after data migration could actually be greater than the number of rows in the table00:08
hemanthmit may confuse operators is all00:08
hemanthmlike how can you migrate 1000 rows when there are only 900 rows in the images table00:09
rosmaitayes, that will provoke some questions00:09
hemanthmwe just double count some rows because of the way we set all non-public images to private first00:09
hemanthmand then set shared00:10
rosmaitayeah, we'll have to think of a nice way to explain that00:11
rosmaitagoing to grab some dinner ... bbiab00:11
hemanthmack, enjoy your dinner00:11
openstackgerritHemanth Makkapati proposed openstack/glance: Add expand/migrate/contract migrations for CI
openstackgerritHemanth Makkapati proposed openstack/glance: Add expand/migrate/contract migrations for CI
dharinichemanthm: I think we should prevent that overlap00:23
dharinicthe count returned for shared must be reassigned as private_count - shared_count00:28
dharinicand then return migrated rowcount00:28
stevellehemanthm: I got nothing but nits here00:29
dharinicor just count only twice. cos the third one is a subset of the second00:34
stevelleI don't think doing calculations makes it any more helpful. the sum is the "number of changes performed" instead of "rows updated" and then it becomes clearer00:35
dharinicHmmm, but as an operator I would expect a result for "number of rows updated".00:36
hemanthmagree with stevelle00:36
hemanthmit depends on how we interpret what's returned from data migrations00:36
dharinicDont see why an operator would care about "number of changes" made00:36
hemanthmdharinic: getting into those calculations is probably okay for this migrations00:37
stevelleIt is unclear to me how number of "rows" is meaningful.00:37
hemanthmbut we don't know if that'd be possible for all migrations00:37
dharinicyep, i realize that.00:38
hemanthmstevelle: feel free to post those nits, I'm waiting on pushing a new PS anyway00:38
stevelleif rows is meaningful, we just do a count(*) and return that number.00:38
stevelleworking on them00:38
hemanthmoh well, I did push a new PS, but I can do another one too00:38
*** ducttape_ has quit IRC00:39
*** ducttape_ has joined #openstack-glance00:39
hemanthmhmmm strange .. I saw gate running on the 2nd patch. Now it says queued again00:40
*** ducttape_ has quit IRC00:41
*** _ducttape_ has joined #openstack-glance00:41
*** _ducttape_ has quit IRC00:42
*** ducttape_ has joined #openstack-glance00:42
stevelleI see is running00:50
stevelle3 checks left to go00:50
dharinicjenkins passes on patch 300:52
hemanthmIt started again after being queued00:53
hemanthmwell, unless I'm hallucinating00:53
*** openstackgerrit has quit IRC01:02
stevellehemanthm: no objections over patch 4, if you're going to do another go for it.01:26
hemanthmstevelle: nothing pending, I didn't realize I already pushed a new PS01:27
stevellehemanthm: wasn't sure if you were going to for
hemanthmstevelle: I did push a new PS for that01:28
hemanthmit is PS #1501:29
stevellenot sure why I'm not seeing it01:29
stevellebombs away.01:31
hemanthmdharinic: are these your first set of +2s?01:32
rosmaitathanks, folks ... lot of work by hemanthm stevelle dharinic and alex_bash01:33
dharinichemanthm: First major +2s01:33
hemanthmah, ok01:34
dharinicI had +2'd 2 small changes earlier01:34
dharinicthanks rosmaita :)01:34
dhariniclot of great work by hemanthm :)01:34
rosmaitawhile we're watching the gate, everyone take a look at the release notes:
hemanthmIt was a lot of work for sure but it is a lot more work to review. Thanks a ton to stevelle, dharinic, rosmaita, nikhil and sigmavirus!01:35
stevelleI'm just glad we ended up having as many eyes on it as we did01:38
hemanthmrosmaita: that's two out of three priorities delivered?01:40
hemanthmor say 1.5 because we didn't finish rolling upgrades01:40
rosmaitawell, 2.5 out of 401:40
rosmaita(community goal, too)01:40
hemanthmah, right01:40
rosmaitai'd even go so far as 2.75 out of 401:41
dharinicrosmaita: the database section you have referenced in the release notes.. would we updating that as well?01:41
rosmaitadharinic: yes, we will have to01:41
rosmaitaif we add a rolling upgrad page, we'll put a link to it on the db page01:42
hemanthmyeah, I will update the dev docs01:42
rosmaitathe advantage of working from the developer docs is that the link in the release notes will be to master01:42
dharinicHmmm. The release note points to something that is currently not updated. Is that fine?01:42
rosmaitait will be updated when the changes merge01:42
dharinicoh cool01:42
rosmaitathat's the db.rst file that was in patch 3 i think01:43
rosmaitabut it will need some more work01:43
rosmaitabut the key thing is, we can merge that whenever we feel like it01:43
*** jose-phillips has quit IRC01:44
*** mtanino has quit IRC01:53
*** openstackgerrit has joined #openstack-glance01:54
openstackgerritFei Long Wang proposed openstack/glance_store master: Replace image.stat() with image.size() call
*** jdillaman has quit IRC02:04
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
*** zhurong has joined #openstack-glance02:08
rosmaitadharinic: sorry, i missed your comments while i was updating that patch02:17
*** jdillaman has joined #openstack-glance02:17
dharinicNo problem rosmaita. They were just some minor suggestions. Anyway we point to the doc02:18
rosmaitadid you mean highlight in italics, or bold?02:18
hemanthmI'm seeing the same thing again ... gate is running twice02:18
*** cdelatte has quit IRC02:19
dharinicback ticks02:19
rosmaitathose are usually reserved for identifiers, though02:19
dharinicActually, any highlight works.02:19
* dharinic needs to learn more about documenting02:20
rosmaitathe two items in the prelude have italics in them, i could see doing that02:20
*** zhurong has quit IRC02:20
dharinicyeah. any form of differentiating them from the other text.. i guess02:21
dharinichemanthm: I think it is running now02:22
rosmaitai'll do that, we can see how it looks ... the manage commands, though, i think shoudl be in the docs, not here02:22
dharinicyep. Sure.02:22
hemanthmdharinic: yeah, but I definitely saw the neutorn jobs almost finish on the 2nd patch02:23
hemanthmthey have started again it looks like02:23
rosmaitai was kind of surprised how fast the first patch merged02:24
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
rosmaitastevelle: flwang followed your hacking suggestion for
flwangrosmaita: thanks for highlight this02:42
flwanganybody know how to check why a test is skipped when using tox? thanks02:43
hemanthmStepping away to get dinner, be back in an hour02:44
*** hemanthm is now known as hemanthm|afk02:44
dharinicflwang: Not sure. But sometimes due to lack of env support in your machine/vm, few tests get skipped.02:51
dharinicby tox02:51
flwangdharinic: yep, that's one of the reasons, but the weird thing is nosetests works fine02:52
dharinicoh. weird. Any specific test thats getting skipped?02:57
*** bkopilov has quit IRC03:02
flwanghah, it's a zaqar test, not glance03:04
flwangso i'm just trying to figure out how to debug it03:04
dharinicAha, okay.03:07
flwangdharinic: but thanks a lot03:07
dharinicflwang: No problem :)03:09
*** mtanino has joined #openstack-glance03:20
*** mtanino has quit IRC03:24
stevellelooks like the gate wedged on post-processing a sporadic ceph failure in 397409, but recheck is running03:34
*** hemanthm|afk is now known as hemanthm03:40
hemanthmoh well03:40
hemanthmdoes the gate have to run again on #3 and #4?03:41
openstackgerritFei Long Wang proposed openstack/glance_store master: Replace image.stat() with image.size() call
*** nicolasbock has quit IRC04:01
*** bkopilov has joined #openstack-glance04:19
stevelleit was cancelled for them, but yes after 2 merges04:24
stevellejust restarted the +1 verify check on 3 and 404:28
hemanthmstevelle: why not reverify?04:30
stevellethey are synonyms.04:30
hemanthmI thought recheck won't merge but reverify does04:30
hemanthmnot sure04:31
stevelleafaik they actually behave the same04:31
*** abhishek_k has joined #openstack-glance04:35
*** abhishekk has quit IRC04:35
*** ducttape_ has quit IRC04:36
*** udesale has joined #openstack-glance04:36
*** pdeore has joined #openstack-glance04:47
*** adisky_ has joined #openstack-glance04:50
hemanthmstevelle: just noticed that you removed +W from #3 and #4. Is that by accident?05:06
stevellenope, will reapply to ensure the gate attempts merge05:07
stevelle#2 just finished the last check for +1, should get the logs up soon and start it's +205:08
*** ratailor has joined #openstack-glance05:08
hemanthmok, thanks05:09
*** bkopilov_ has joined #openstack-glance05:25
*** groen692 has joined #openstack-glance05:35
*** ducttape_ has joined #openstack-glance05:37
*** groen692 has quit IRC05:42
*** groen692 has joined #openstack-glance05:42
*** ducttape_ has quit IRC05:42
*** markvoelker has joined #openstack-glance06:10
*** markvoelker_ has quit IRC06:12
*** mnaser has quit IRC06:12
*** abhishek_k has quit IRC06:13
*** abhishek_k has joined #openstack-glance06:13
*** groen692 has quit IRC06:27
*** groen692 has joined #openstack-glance06:29
*** mnaser has joined #openstack-glance06:35
*** groen692 has quit IRC06:42
*** groen692 has joined #openstack-glance06:44
*** clayton has quit IRC06:46
*** groen692 has quit IRC06:50
*** groen692 has joined #openstack-glance06:50
*** pdeore has quit IRC06:57
*** groen692 has quit IRC07:00
*** pdeore has joined #openstack-glance07:04
*** rcernin has joined #openstack-glance07:06
*** tshefi has joined #openstack-glance07:07
*** ducttape_ has joined #openstack-glance07:08
*** ducttape_ has quit IRC07:13
*** zhurong has joined #openstack-glance07:13
*** tesseract has joined #openstack-glance07:16
*** groen692 has joined #openstack-glance07:21
*** e0ne has joined #openstack-glance07:28
*** groen692 has quit IRC07:33
*** e0ne has quit IRC07:33
*** mfedosin has joined #openstack-glance07:34
*** e0ne has joined #openstack-glance07:38
*** pcaruana has joined #openstack-glance07:39
*** e0ne has quit IRC07:42
*** pcaruana has quit IRC07:42
*** pcaruana has joined #openstack-glance07:43
*** e0ne has joined #openstack-glance07:46
*** groen692 has joined #openstack-glance07:50
*** e0ne has quit IRC07:53
*** e0ne has joined #openstack-glance07:57
*** e0ne has quit IRC08:01
*** e0ne has joined #openstack-glance08:06
*** e0ne has quit IRC08:11
*** e0ne has joined #openstack-glance08:16
*** e0ne has quit IRC08:21
*** e0ne has joined #openstack-glance08:26
*** zhurong has quit IRC08:28
*** rcernin has quit IRC08:28
*** ezoszed has joined #openstack-glance08:29
*** ezoszed has quit IRC08:29
*** ezoszed has joined #openstack-glance08:31
*** e0ne has quit IRC08:34
*** ducttape_ has joined #openstack-glance08:38
*** e0ne has joined #openstack-glance08:38
*** ducttape_ has quit IRC08:43
*** e0ne has quit IRC08:50
*** rcernin has joined #openstack-glance08:52
*** e0ne has joined #openstack-glance08:54
*** zzzeek has quit IRC09:00
*** zzzeek has joined #openstack-glance09:01
BobBallrosmaita: No worries - thanks.09:02
BobBallrosmaita: Looks like the failure might have been Cinder breaking support for Ubuntu 14.0409:02
*** e0ne has quit IRC09:02
*** ezoszed has quit IRC09:09
*** ezoszed has joined #openstack-glance09:17
openstackgerritDanny Al-Gaaf proposed openstack/glance master: Adds purge command to glancemanage man page
stevelle :(10:09
*** ducttape_ has joined #openstack-glance10:09
*** ducttape_ has quit IRC10:14
stevelle may resolve that, will have to see10:19
*** udesale has quit IRC10:24
*** dalgaaf has quit IRC10:43
*** dalgaaf has joined #openstack-glance10:49
*** aarefiev_afk is now known as aarefiev11:08
*** ducttape_ has joined #openstack-glance11:10
*** ducttape_ has quit IRC11:14
*** nicolasbock has joined #openstack-glance11:18
*** ducttape_ has joined #openstack-glance11:30
*** groen692 has quit IRC11:35
*** groen692 has joined #openstack-glance11:47
*** catintheroof has joined #openstack-glance11:47
*** groen692 has quit IRC11:52
*** ducttape_ has quit IRC11:57
*** bkopilov_ has quit IRC11:57
*** bkopilov has quit IRC11:58
*** ducttape_ has joined #openstack-glance12:00
*** e0ne has joined #openstack-glance12:00
*** ratailor has quit IRC12:02
*** openstackgerrit has quit IRC12:03
*** clayton has joined #openstack-glance12:13
*** ducttape_ has quit IRC12:22
*** ducttape_ has joined #openstack-glance12:42
*** ducttape_ has quit IRC12:53
*** mvk has quit IRC12:55
*** gcb has joined #openstack-glance13:00
*** ducttape_ has joined #openstack-glance13:02
*** openstackgerrit has joined #openstack-glance13:09
openstackgerritMerged openstack/glance master: Refactor tests to use Alembic to run migrations
*** dalgaaf has quit IRC13:20
*** dalgaaf has joined #openstack-glance13:27
*** mvk has joined #openstack-glance13:29
*** abashmak_ is now known as alex_bash13:47
*** zhurong has joined #openstack-glance13:48
*** ducttape_ has quit IRC13:57
rosmaitaCourtesy meeting reminder on #openstack-meeting-4: ativelkov, cpallares, flaper87, flwang1, hemanthm, jokke_, kragniz, lakshmiS, mclaren, mfedosin, nikhil_k, Nikolay_St, Olena, pennerc, rosmaita, sigmavirus24, sabari, TravT, ajayaa, GB21, bpoulos, harshs, abhishek, bunting, dshakhray, wxy, dhellmann, kairat, aavraham, alex_bash13:59
*** gcb has quit IRC14:05
*** zhurong has quit IRC14:06
*** ducttape_ has joined #openstack-glance14:16
*** zhurong has joined #openstack-glance14:17
*** bkopilov_ has joined #openstack-glance14:19
*** bkopilov has joined #openstack-glance14:19
*** ducttape_ has quit IRC14:27
*** udesale has joined #openstack-glance14:29
sigmavirushemanthm: SSH timeout should be safe to recheck:
sigmavirushemanthm: also, us old fogies should unlearn "reverify" since everything responds to "recheck" and "reverify" never stopped the patch from going back through the check queue to the gate queue14:37
hemanthmsigmavirus: yeah, it's the same failure that keeps showing up14:37
hemanthmsigmavirus: ack14:37
*** zzzeek has quit IRC14:40
*** zzzeek has joined #openstack-glance14:40
*** pdeore has quit IRC14:49
*** TravT has joined #openstack-glance14:59
*** udesale has quit IRC15:00
dharinicrosmaita: Could use some advise on the translation issue. Since this is cherry picked, i do not get why it would fail on a releasenote.15:01
rosmaitadharinic: looking in stable/newton, it looks like translations were imported fairly recently15:02
rosmaitabut the was not changed15:02
dharinicMy first recheck for this issue was on Jan 5 rosmaita15:02
rosmaita"recent" == 7 months ago15:02
dharinicHaha. okay15:03 dates back to a year ago15:03
dhariniccool, so this calls for a change to
rosmaitai think so ... maybe the thing to do is see if dhellmann is in #openstack-infra and see if there's something about reno and translations15:04
dharinicokay. Will check. Thanks rosmaita15:04
rosmaitabut the short answer is, i think you're going to have to modify conf.py15:06
rosmaitabut i don't really understand why ... it would be great if you could find out15:06
*** zhurong has quit IRC15:09
*** mtanino has joined #openstack-glance15:13
dharinicwill look rosmaita15:13
*** ducttape_ has joined #openstack-glance15:14
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
rosmaitadharinic: hemanthm: stevelle: could use a sanity check on ^^15:15
rosmaitaall i can say is, we have some seriously long release notes!15:16
*** johndperkins has quit IRC15:17
hemanthmbetter long than short I guess15:18
*** TravT has quit IRC15:19
*** TravT has joined #openstack-glance15:19
*** gabor_antal_ has joined #openstack-glance15:32
*** gabor_antal has quit IRC15:32
*** konan_abhi has joined #openstack-glance15:34
*** abhishek_k has quit IRC15:35
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
*** thiagolib has quit IRC15:38
*** ezoszed has quit IRC15:44
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
alex_bashdharinic: congrats on making core!!!15:50
*** e0ne has quit IRC15:53
rosmaitasigmavirus: sigmaroosavirus it is!15:54
sigmavirusrosmaita: glad you liked that =P15:55
rosmaitai may have to refer to you as "teddy" occasionally, though15:56
*** burgerk has joined #openstack-glance15:59
*** _ducttape_ has joined #openstack-glance16:00
*** _ducttape_ has quit IRC16:01
*** ducttape_ has quit IRC16:03
*** ducttape_ has joined #openstack-glance16:06
stevellepoint of interest: should clean up some of the gate failures we're being hit by so things can merge faster16:06
hemanthmthat one is just ahead of us in the gate16:08
dharinicThank you alex_bash :)16:08
stevellemaybe don't workflow or recheck until it merges, hemanthm16:09
hemanthmstevelle: +116:10
*** mfedosin has quit IRC16:14
*** kberger has joined #openstack-glance16:25
*** rcernin has quit IRC16:29
*** ducttape_ has quit IRC16:38
*** ducttape_ has joined #openstack-glance16:39
*** tesseract has quit IRC16:41
openstackgerritMerged openstack/glance master: Add expand/migrate/contract commands to glance-manage CLI
hemanthmthis is ready for +W16:52
rosmaitai'll do it16:53
rosmaitai was going to add, "and may God have mercy on our souls" to my +A comment, but thought that would be overdoing it16:57
hemanthmrosmaita: it's your fault now if rolling upgrades break, you didn't ask for mercy16:58
rosmaitai just got back from the api-wg meeting ... really don't want to argue theology now!16:58
rosmaitatrue story though ... at a previous job where i was tech lead on a project, i used to go donate blood the day before a release, because i felt that the software required a blood sacrifice17:00
hemanthmIf that was true, I don't know what sacrifice OpenStack would need17:01
alex_bashliver and at least one of the kidneys17:01
rosmaitaalex_bash: problem is, eventually you run out of those!17:02
alex_bashthat's why we need fresh contributors all the time17:03
*** hoonetorg has quit IRC17:03
alex_bashshhh, don't tell the newbies their true purpose17:03
rosmaitareally fresh!17:03
stevelleseems legit17:03
stevelleso, lets wrap up the work on and get it into the gate17:04
*** hoonetorg has joined #openstack-glance17:04
stevellewith the nova fix in we should be pretty safe, not needing to do 1 at a time17:04
dharinicI had a question on the releasenote rosmaita17:05
rosmaitacool ... thanks for watching all this stevelle17:05
rosmaitayour eyes must be glazed over17:05
rosmaitayou've been at it for like 48 hours17:05
rosmaitadharinic: sure17:05
rosmaitayes, that is a good question17:05
dharinicI think stevelle mentioned it yesterday too. We need not really be talking about scripts to operators..17:06
rosmaitawell, i wonder whether operators modify scripts17:06
rosmaitalike add extra indexes and stuff17:06
dharinicLess likely, but possible17:06
rosmaitaalso, probably before doing an upgrade, you want to see what you're getting into17:07
dharinicIf everybody thinks, it is okay to have all that information in there, I am cool.17:07
stevelleI don't feel strongly, just wanted to get the thinking out loud on the point17:07
dharinicJust felt that it was too in-detail17:07
rosmaitai was wondering about that (the too much detail, i mean)17:07
stevelleand the audience includes packagers and operators17:07
rosmaitayeah, my reason for putting it in was that i figure packagers are going to include both17:08
rosmaitasince we didnt' remove the "old" one17:08
rosmaitadharinic: but i'm still not convinced it's the right move17:08
rosmaitabummer that nikhil isn't here, he has spent a lot of time talking to operators17:09
nikhilI'm partially present17:09
* nikhil has no context yet17:10
rosmaitanikhil: ^^ will give you some context17:10
hemanthmIsn't too much detail better than missing detail?17:10
hemanthmIf we are not sure I mean17:10
rosmaitai'm not sure ... we dont' want to encourage people to mess around17:11
rosmaitabut we dont' want to hide anything, either17:11
rosmaita(and we can't actually hide anything, anyway)17:11
hemanthmWell, all the information is going to be in the dev docs any way17:12
hemanthmand there is a link to dev docs in the release notes17:12
hemanthmwe can't stop people from messing around17:12
hemanthmMy only worry is repetition17:13
rosmaitaright ... i am going to put up a dev docs patch later today17:13
rosmaitaso i could pull that stuff out of hte relnotes and use it in the devdocs17:13
rosmaitai guess the advantage of the detail is that it shows that we mean business17:14
rosmaitai.e., this is a change that needs to be paid attention to17:14
rosmaitabecause the commands for the manage cli take different values for the versions17:15
hemanthmI don't have a strong opinion, I'm good with what it is now17:17
hemanthmI don't think the extra detail hurts17:17
*** mtanino_ has joined #openstack-glance17:18
dharinicyeah, as long as it is not too much of repeating information, it is okay i guess17:19
rosmaitathanks for flagging this, dharinic , because it's a big pain to backfix a relnote, so it would be good to be convinced this is what we want before we merge it17:19
*** mtanino__ has joined #openstack-glance17:19
*** mtanino has quit IRC17:19
dharinicsure rosmaita. We can merge it when we are sure we dont have information repeating in both the rel note and the doc17:20
rosmaitawell, there will definitely be some repeating info17:20
rosmaitathe locations and stuff for sure17:20
rosmaitathose have to be in teh dev docs17:20
rosmaitapretty much the entire thing, actually, just with more detail17:20
rosmaitai thought it was key to mention that there are 2 script directories now, but only one of them is actionable17:21
rosmaitaso then needed to say which is the "real" one, and what the point of having the other one around is17:22
*** mtanino_ has quit IRC17:22
rosmaitai'll be curious what nikhil says when he's had a chance to look it over17:22
nikhilrosmaita: looking now17:29
*** hoonetorg has quit IRC17:38
*** chlong has joined #openstack-glance17:39
*** ducttape_ has quit IRC17:39
*** hoonetorg has joined #openstack-glance17:41
*** aarefiev is now known as aarefiev_afk17:44
*** hoonetorg has quit IRC17:47
nikhilrosmaita: commented17:48
rosmaitanikhil: what do you think about removing the "old" migration scripts before RC-1 ?17:54
nikhilrosmaita: you mean the code?17:55
nikhilor reference in the docs?17:55
rosmaitaout of the code tree17:55
*** jose-phillips has joined #openstack-glance17:56
sigmavirusrosmaita: ping me when the last two rolling upgrade patches merge?17:56
hemanthmsigmavirus: 3rd one merged17:56
* sigmavirus thought 3/5 were merged over an hour ago17:56
hemanthm4th is in the gate and almost done17:57
rosmaita5th is being revised, but it's docs only, should go pretty fast17:57
nikhilrosmaita: hmm, I thought this feature is experimental?17:57
rosmaitaonly the rolling upgrade part17:58
rosmaitathe E-M-C migration17:58
rosmaitaotherwise, you can just do a regular upgrade17:58
rosmaitabut the regualr one uses alembic17:58
sigmavirusrosmaita: good use of hyphens to avoid copyright issues17:58
rosmaitasigmavirus: :)17:58
sigmavirusregular == cold, yes?17:58
nikhilrosmaita: noted17:58
*** hoonetorg has joined #openstack-glance17:59
nikhilrosmaita: please don't get me wrong and I don't believe we've reason to doubt: but if a few come back shouting I think we can have a backup plan17:59
rosmaitaso you are saying keep the old scripts in?17:59
sigmavirusnikhil: how do you mean?17:59
nikhilrosmaita: we can mark them deprecated (I think)17:59
hemanthmI think we should keep old ones around for a bit18:00
rosmaitawell, they won't run with glance-manage anymore18:00
hemanthmMaybe remove them in Pike18:00
nikhilthough that won't have any technical significance18:00
hemanthmHow do we mark migrations deprecated?18:00
nikhilyeah, we can't show them in the output18:01
nikhilso the deprecation has less meaning18:01
hemanthmThere is no execution path for them18:01
rosmaitawell, its worse than that ... the same commands now assume alembic scripts18:01
rosmaitawhat hemanthm said18:01
hemanthmUnless someone explicitly installs and invokes sqla-migrate18:01
nikhilbut the feeling a devops person gets when he/she wakes up a fine morning to see them deleted18:02
rosmaitasure, but they are replaced with brand-spanking-new ones!18:02
rosmaitaso ... that's why i went into all that detail in the release note18:02
nikhilit's a bad situation to make downstream folks with custom patches really really annoyed18:03
rosmaitaso if someone looks, they won't be surprised/confused18:03
nikhilit's a hit and miss18:03
hemanthmThey would be annoyed anyway18:03
hemanthmIf they had their custom migrations, they'd have to switch to alembic too18:03
nikhilor revert alembic commits18:04
rosmaitathat's the advantage of both versions, i guess, they can see how they will have to change their patches18:04
nikhilin the later case it's a easier devops experience18:04
hemanthmI want to keep them around for the reason rosmaita said above18:05
nikhilI don't believe all the problems are independent18:05
rosmaitawell, for the convenience of operators, it just so happens that hte alembic commits are all stacked up right at the top!18:05
nikhilor any arguments for that matter18:05
nikhilthings are interconnected18:05
rosmaitaok, so the consensus is: (a) leave the old scripts in18:06
nikhilrosmaita: yep, if people are annoyed it will be a hard sell18:06
rosmaitahow about (b) remove the locations from the release note <-- i dont like this so much18:06
nikhil(a) is for developers/devops18:07
nikhil(b) is for operators to not get scared18:07
hemanthm(b) is just about how to present it18:08
hemanthmIt'll be there in the link anyway18:08
hemanthmlink to dev docs18:08
nikhil+1 on how to present it18:09
rosmaitaok, i'll do a quick revision and we can see how it looks18:09
*** e0ne has joined #openstack-glance18:19
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
rosmaitanikhil: hemanthm: stevelle: dharinic: sigmavirus: ^^18:23
openstackgerritMerged openstack/glance master: Add expand/migrate/contract migrations for CI
rosmaitasigmavirus: ^^ that was #4, now just the release notes18:29
sigmavirusrosmaita: so18:29
sigmavirusif you rebase your releasenotes patch on top of current master and there are no objects, I can submit your SHA sicne it'll merge without a emrge commit18:30
sigmavirusif you don't want to do that, then we can just wait for yours to merge to submit the release request18:30
rosmaitaok ... need comments on this version first18:32
*** dharinic is now known as dharinic|lunch18:34
sigmavirusgeeze rosmaita you're so needy ;P18:35
*** mvk has quit IRC18:35
rosmaitawhat i need is some lunch ... but will get this done first18:37
hemanthmrosmaita: go ahead, I can push up a patch if further changes are needed18:38
* sigmavirus aims to make review of this patch as humorous as possible18:39
rosmaitai'll make sigmavirus 's changes and then go18:39
sigmavirusI'm good after those changes are made18:39
rosmaitai'm glad someone appreciates my experimental prose18:39
sigmavirusYou'll have my Roose+218:39
*** ducttape_ has joined #openstack-glance18:40
sigmavirusalso never forget that wen I started calling kragniz "kraggy" he started calling me "siggy"18:40
sigmavirusLong live Kraggy18:40
nikhilrosmaita: you've my comments too18:42
*** ducttape_ has quit IRC18:44
rosmaitanikhil: ty18:46
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
hemanthmrosmaita: "for" is repeated on L3418:57
* nikhil bbiab19:00
openstackgerritBrian Rosmaita proposed openstack/glance master: Alembic migrations/rolling upgrades release note
rosmaitai hope that's the last one ^^19:03
* rosmaita bbiab19:08
*** dharinic|lunch is now known as dharinic19:08
stevellesomewhat surprised nobody followed nikhil's statement with "and my axe!"19:09
hemanthmI get this reference at least19:16
nikhilstevelle: I dont19:26
stevellefair enough. just a lord of the rings thing.19:27
*** ducttape_ has joined #openstack-glance19:32
rosmaitasigmavirus: your move ... want a new ps?19:32
rosmaitawho'd have thought we'd be talking about the present perfect simple tense on a review? gotta love working on glance!19:35
*** TravT has quit IRC19:37
sigmavirusrosmaita: nah, I'm fine19:37
*** mtanino__ has quit IRC19:39
hemanthmany final reservations anyone? If not, will +W19:39
rosmaitasigroosavirus beat you to it19:40
hemanthmoh, nevermind .. Roosevirus to the rescue19:40
sigmaviruset. all ^20:03
sigmavirusrosmaita: just fwiw, a docs person agrees with me on your release notes change =P20:03
sigmavirussaid docs person will remain unnamed20:03
rosmaitai can live with that20:17
rosmaitayou probably noticed that my sentences aren't the shortest, either20:18
nikhilrosmaita: congrats on getting this out20:19
rosmaitabut since we use python, i've been able to remove the semicolon key from my keyboard20:19
rosmaitanikhil: thanks for your help, some very timely reviews20:19
nikhilrosmaita: ;)20:20
nikhilrosmaita: well, python does have semicolon20:20
nikhilrosmaita: but it also has zen of python20:20
rosmaitai thought we had a hacking rule against that :)20:21
* nikhil checks20:21
sigmavirusrosmaita: we do20:21
sigmavirusit's not actually hacking specific20:21
sigmavirusit's pep8 specific20:21
sigmavirusGlance RC-1 has been approved!20:22
rosmaitaman, i was just looking over the hash20:22
sigmavirus#success Glance merged Experimental Rolling Upgrades for RC-120:23
openstackstatussigmavirus: Added success to Success page20:23
sigmavirus(we should use that more often)20:23
nikhilrosmaita: sigmavirus : congrats on getting RC1 out20:24
rosmaitai had no idea20:24
nikhilsigmavirus: couldn't find the rule yet20:24
sigmavirusrosmaita: it's nice20:24
sigmavirusnikhil: it's a rule about multiple statements on one line20:24
nikhilrather pep8 has another pet peeve
sigmavirusI don't think it specifically checks for `;` only20:24
nikhilsigmavirus: aha20:24
nikhilit's weird20:25
sigmaviruspeople tend to only use ;s in python to have more than one statement on a line20:25
sigmaviruswhat do you mean?20:25
sigmavirusOh yeah20:25
nikhilhow that's a inconsistent cyclic dep20:25
sigmavirusSome people love the whitespace inside parens20:25
sigmavirusI never understood that20:25
sigmavirusI've never seen that in the wild in any language either20:25
nikhilin that rule they mention semicolon20:25
sigmavirusmultiple expressions on a single line are *okay*20:26
sigmavirusBut they prefer you don't20:26
sigmavirusBut ifyou have to20:26
sigmavirusDon't do it that way20:26
nikhildo it only if you don't :P :P20:26
nikhil(don't have to)20:27
sigmavirusGuido, et. all, have been nagged to clarify some things over the years in that doc20:27
*** mfedosin has joined #openstack-glance20:27
sigmavirusbecause some people use it as a book that one can never violate and bash each other's heads in with it20:28
sigmavirusso some silly clarifications have landed20:28
nikhil(don't doubt it)20:29
*** mvk has joined #openstack-glance20:38
*** TravT has joined #openstack-glance20:38
*** TravT has quit IRC20:39
rosmaitasigmavirus: i forgot to regenerate the config files20:40
*** harlowja has quit IRC21:00
*** _ducttape_ has joined #openstack-glance21:06
*** ducttape_ has quit IRC21:10
*** mfedosin has quit IRC21:14
*** catintheroof has quit IRC21:14
*** catintheroof has joined #openstack-glance21:15
sigmavirusrosmaita: poop21:19
*** catintheroof has quit IRC21:19
*** _ducttape_ has quit IRC21:23
*** ducttape_ has joined #openstack-glance21:24
rosmaitasigmavirus: ploop21:43
rosmaitarelease notes still haven't merged21:43
stevellegood call on the use of success tracker sigmavirus21:56
stevelleI forget abt it21:56
*** flwang has left #openstack-glance22:03
dhellmannsigmavirus, rosmaita : do you have any release blockers in progress that should hold up the glance rc1 tag?22:03
dhellmannit would be better to wait than to rush rc1 and have an rc2 follow on quickly after22:04
dhellmannthat tends to train users to ignore release candidates22:04
rosmaitadhellmann: yes, i think i need to merge a config file update22:04
dhellmannrosmaita : ok, maybe mark WIP until you're ready?22:04
dhellmannoh, I see your -1 now22:04
dhellmannmy dashboard hides the -1 when there is a +2 :-/22:05
rosmaitayeah, i don't have workflow powers22:05
dhellmannnp, I'll remove my +2 for now22:05
rosmaitathat change you're waiting on still hasn't merged, not sure what's up wtih that22:05
rosmaitadhellmann: thanks22:05
rosmaitai will update that patch soon as i have a new hash22:06
*** harlowja has joined #openstack-glance22:25
*** ducttape_ has quit IRC22:26
*** ducttape_ has joined #openstack-glance22:31
*** harlowja has quit IRC22:32
*** harlowja has joined #openstack-glance22:33
*** ducttape_ has quit IRC22:43
*** mtanino has joined #openstack-glance22:51
-openstackstatus- NOTICE: Restarting gerrit due to performance problems22:51
* stevelle is glad glance is out of the gate right now22:51
rosmaitaactually, release notes are still in there22:51
stevelleah, thought we had that through22:52
rosmaitano, and it looks like that thing hemanthm noticed yesterday when things are going through twice22:53
*** burgerk_ has joined #openstack-glance23:00
stevellehmm, well it looks like the job is done, acting like the logstack queue having trouble or something23:01
*** burgerk has quit IRC23:03
*** burgerk_ has quit IRC23:04
*** tshefi has quit IRC23:05
openstackgerritMerged openstack/glance master: Alembic migrations/rolling upgrades release note
openstackgerritBrian Rosmaita proposed openstack/glance master: Refresh config files for Ocata RC-1
openstackgerritBrian Rosmaita proposed openstack/glance master: Refresh config files for Ocata RC-1

Generated by 2.14.0 by Marius Gedminas - find it at!