| *** openstack has joined #openstack-lbaas | 00:11 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Use rfc3986 module for URL validations https://review.openstack.org/290918 | 00:11 |
|---|---|---|
| *** ducttape_ has quit IRC | 00:12 | |
| *** openstackgerrit has quit IRC | 00:17 | |
| *** bank has quit IRC | 00:18 | |
| *** openstackgerrit has joined #openstack-lbaas | 00:18 | |
| *** bank has joined #openstack-lbaas | 00:20 | |
| *** Paco_ has joined #openstack-lbaas | 00:35 | |
| *** purandar_k has joined #openstack-lbaas | 00:46 | |
| *** clduser_ has joined #openstack-lbaas | 00:48 | |
| *** ptoohill has quit IRC | 00:49 | |
| *** Purandar has quit IRC | 00:49 | |
| *** Kiall has quit IRC | 00:49 | |
| *** clduser has quit IRC | 00:49 | |
| *** Kiall has joined #openstack-lbaas | 00:49 | |
| *** ptoohill has joined #openstack-lbaas | 00:49 | |
| *** chenli has joined #openstack-lbaas | 00:50 | |
| *** cody-somerville has quit IRC | 00:51 | |
| *** numans has joined #openstack-lbaas | 00:51 | |
| *** numans has quit IRC | 00:58 | |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Add the Angular LBaaS V2 'Edit Pool' Workflow https://review.openstack.org/289453 | 01:05 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Add health monitor delete action https://review.openstack.org/287503 | 01:08 |
| *** ducttape_ has joined #openstack-lbaas | 01:08 | |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Add create health monitor action https://review.openstack.org/288559 | 01:08 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Add edit health monitor action https://review.openstack.org/288607 | 01:08 |
| *** ducttape_ has quit IRC | 01:14 | |
| *** ducttape_ has joined #openstack-lbaas | 01:15 | |
| *** cody-somerville has joined #openstack-lbaas | 01:22 | |
| *** neelashah has joined #openstack-lbaas | 01:27 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Use consistent type checking of project_id in API https://review.openstack.org/290936 | 01:32 |
| sbalukoff | johnsom, rm_work, blogan: I think this ^^^ is a critical bugfix. | 01:34 |
| sbalukoff | This is the corresponding bug report: | 01:34 |
| sbalukoff | https://bugs.launchpad.net/octavia/+bug/1555401 | 01:34 |
| openstack | Launchpad bug 1555401 in octavia "Incorrect project_id type in API results in erratic behavior" [Undecided,In progress] - Assigned to Stephen Balukoff (sbalukoff) | 01:34 |
| johnsom | Ok, will look in a bit | 01:37 |
| johnsom | Thanks! | 01:37 |
| *** minwang2 has quit IRC | 01:45 | |
| *** ducttape_ has quit IRC | 01:45 | |
| *** armax has quit IRC | 01:46 | |
| *** cody-somerville has quit IRC | 01:48 | |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Add action for updating pool member weight https://review.openstack.org/288004 | 01:48 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Add trailing slash to all URLs calling the REST API https://review.openstack.org/290895 | 01:49 |
| *** fawadk has quit IRC | 01:49 | |
| johnsom | sbalukoff Umm, we have a conflict: https://review.openstack.org/#/c/282574/ | 01:58 |
| johnsom | I, as you know, advocate the uuid types, but there is a reason for those to not be uuid | 01:59 |
| sbalukoff | HAHA | 01:59 |
| sbalukoff | Is the reason because the project_id as passed in by Neutron-LBaaS doesn't contain hyphens? | 02:00 |
| johnsom | I vaguely remember that some operator was using non-uuid project_ids | 02:01 |
| sbalukoff | Ugh. | 02:02 |
| sbalukoff | Ok... | 02:02 |
| johnsom | Check with ptoohill and blogan | 02:02 |
| sbalukoff | I'll update my patch so that at least we're treating it consistently across the board. I'm not sure how the regression crept in. | 02:02 |
| sbalukoff | We really just need to be consistent. | 02:02 |
| johnsom | Well, probably because I beat on the table that things should be uuid types | 02:02 |
| johnsom | grin | 02:02 |
| *** fawadkhaliq has joined #openstack-lbaas | 02:05 | |
| johnsom | Funny that the other patch was just 19 days ago | 02:05 |
| sbalukoff | Yeah. | 02:05 |
| sbalukoff | It's been a long 19 days. | 02:05 |
| johnsom | Yes, yes it has | 02:06 |
| *** yamamoto_ has joined #openstack-lbaas | 02:10 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Use consistent type checking of project_id in API https://review.openstack.org/290936 | 02:22 |
| sbalukoff | johnsom: There ^^^^. Also, I added unit tests so we don't get another regression any time soon. :P | 02:22 |
| johnsom | I like more tests | 02:27 |
| *** bank has quit IRC | 02:27 | |
| *** kev0 has quit IRC | 02:27 | |
| *** yamamoto_ has quit IRC | 02:49 | |
| *** fawadkhaliq has quit IRC | 02:53 | |
| openstackgerrit | Trevor Vardeman proposed openstack/neutron-lbaas: WIP - Get Me A LB https://review.openstack.org/257201 | 03:03 |
| *** ajmiller has quit IRC | 03:10 | |
| *** ducttape_ has joined #openstack-lbaas | 03:11 | |
| *** ducttape_ has quit IRC | 03:11 | |
| *** madhu_ak has quit IRC | 03:11 | |
| *** Paco_ has quit IRC | 03:15 | |
| *** fawadkhaliq has joined #openstack-lbaas | 03:24 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Fix L7Policy update regression https://review.openstack.org/290963 | 03:28 |
| *** purandar_k has quit IRC | 03:33 | |
| *** fawadkhaliq has quit IRC | 03:34 | |
| *** yamamoto_ has joined #openstack-lbaas | 03:36 | |
| *** Aish has left #openstack-lbaas | 03:38 | |
| *** bank has joined #openstack-lbaas | 03:42 | |
| *** links has joined #openstack-lbaas | 03:50 | |
| *** Purandar has joined #openstack-lbaas | 03:55 | |
| *** woodster_ has quit IRC | 03:57 | |
| *** chenli has quit IRC | 04:00 | |
| blogan | how did that happen | 04:06 |
| blogan | TrevorV!!! | 04:07 |
| blogan | single create call, merge conflict fix fail, glad you added the tests sbalukoff | 04:07 |
| sbalukoff | Eh... I've done worse. | 04:10 |
| sbalukoff | Found a regression after the better l7policy update code. | 04:10 |
| sbalukoff | (Which I *think* I have fully fixed now... just running tox locally and will upload a new patch set.) | 04:10 |
| sbalukoff | Also blogan: | 04:10 |
| blogan | i'm just giving him a hard time, we've all done fun stuff like that | 04:10 |
| sbalukoff | It is a serious pain in the ass that pecan fills in unspecified parameters in PUT requests with 'None' values. | 04:11 |
| sbalukoff | I think there's a lot of work-around code I could eliminate if we could get it to stop doing that. | 04:11 |
| blogan | sbalukoff: its probably WSME that does that | 04:11 |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Fix L7Policy update regression https://review.openstack.org/290963 | 04:11 |
| blogan | sbalukoff: and i think there's a Undefined type? | 04:12 |
| sbalukoff | blogan: Aah... well whoever is to blame, it's annoying! | 04:12 |
| blogan | let me check out the code | 04:12 |
| sbalukoff | I couldn't find much documentation when I looked. :/ | 04:12 |
| blogan | sorry, wsme has an Unset type | 04:12 |
| blogan | are you seeing None on the actual type or after its been converted to a dict? | 04:13 |
| blogan | sbalukoff: https://github.com/openstack/octavia/blob/master/octavia/api/v1/types/base.py#L71 | 04:13 |
| sbalukoff | Yes. | 04:14 |
| sbalukoff | Yeah, that's a problem. | 04:14 |
| sbalukoff | Sometimes it is helpful if the API request explicitly set a given value to None. | 04:14 |
| sbalukoff | Oh, so *we're* the ones doing it... | 04:15 |
| sbalukoff | Why did we do that? | 04:16 |
| fnaval | sbalukoff: i commented on bug 1547609 | 04:16 |
| openstack | bug 1547609 in octavia "Delete Health Monitor makes LB go into ERROR operating status" [Critical,Incomplete] https://launchpad.net/bugs/1547609 - Assigned to Stephen Balukoff (sbalukoff) | 04:16 |
| blogan | sbalukoff: tha tshouldn't change something specified as None | 04:17 |
| sbalukoff | blogan: It changes things that are *not* specified to None. | 04:17 |
| sbalukoff | That's the problem. | 04:17 |
| blogan | on PUTs? | 04:17 |
| sbalukoff | Yes. | 04:18 |
| blogan | the queue handler is doing render_unsets=False | 04:18 |
| sbalukoff | Right. So that gets rid of anything that *was* specified to be None. | 04:18 |
| blogan | how? | 04:19 |
| sbalukoff | Let me check my notes again... | 04:20 |
| blogan | if its not specified WSME will literally set the value of that attribute to Unset, and the to_dict method only changes those attributes with the Unset type if render_unset=TRue, otherwise everything should be the same | 04:20 |
| sbalukoff | Augh, dammit-- I think I might be getting bitten in the ass by the default on that method: render_unsets=True | 04:22 |
| sbalukoff | Fuck. | 04:22 |
| sbalukoff | I could swear this was broken in other ways, though... Hmmm.... | 04:23 |
| blogan | yeah i was just thinking that could throw people for a loop, and i dont remember why i set the default to that, probably bc thats what create needed | 04:23 |
| blogan | sbalukoff: lol i dont doubt that | 04:23 |
| blogan | i like wsme, but sometimes i dont, not a lot of documentation on it | 04:24 |
| sbalukoff | This was us shooting ourselves in the foot | 04:25 |
| sbalukoff | By having a non-obvious option. It is unexpected that an API request that specifically does NOT set certain values should show up in the API controller's post or put methods with the missing values filled in. | 04:25 |
| sbalukoff | Ok, I think I've got three or four patches to re-work now... | 04:26 |
| blogan | we can totally change the default and let the creates specify render_unsets=True, not sure if thats the only reason for it though | 04:27 |
| blogan | though i would argue that it makes sense for creates, obviously for puts it does not | 04:27 |
| sbalukoff | For puts it's actually a very bad thing. | 04:28 |
| sbalukoff | Thankfully, I think only the L7 code is hosed by this-- and I've already got work-arounds (which I can now make much simpler) in place for this. This is because only the L7 code actually does any decent validations. :P | 04:28 |
| openstackgerrit | Michael Johnson proposed openstack/octavia: Fix LB and Listener admin_state Update https://review.openstack.org/290408 | 04:29 |
| blogan | L7 probably has some of hte more complex validation needed too | 04:29 |
| sbalukoff | Screw it. I'm not going to change the default value for that. I'll just fix the API code for L7 stuff. | 04:29 |
| blogan | though validating the subnet_ids will be complex if we ever get that in | 04:29 |
| sbalukoff | (Also need to fix this for the session_persistence update bug-- as that is affected by this.) | 04:30 |
| blogan | sbalukoff: did we start updating the db first before passing to the queue on PUTs? | 04:31 |
| sbalukoff | No. | 04:31 |
| blogan | so i dont see how this would cause problems in the current codebase, unless its only in your review | 04:32 |
| *** ajmiller has joined #openstack-lbaas | 04:33 | |
| blogan | sbalukoff: ah i see your review, at least one of them | 04:35 |
| sbalukoff | Yep. | 04:36 |
| sbalukoff | Marking them WIP so nobody merges them until I have a chance to revisit this. | 04:36 |
| sbalukoff | the session persistence bug fix was a serious pain in the ass, so fixing the work-around (if it needs to be fixed) is going to be a pain as well. | 04:37 |
| sbalukoff | And I've got to rebuild my test environment to do that. | 04:37 |
| *** amotoki has joined #openstack-lbaas | 04:38 | |
| sbalukoff | All because I didn't realize the fucking to_dict() method was doing stupid, unexpected shit in the debug output I was using to troubleshoot (but that the underlying objects were potentially just fine.) | 04:38 |
| sbalukoff | :P | 04:38 |
| blogan | well again i wouldn't say it was stupid, unexpected shit, okay well yes for puts, but for creates it made sense | 04:39 |
| blogan | perhaps if it was a positional arg that have been better, just to force people to see what that arg was | 04:39 |
| sbalukoff | In a lot of other code we have special args that get passed for create methods. | 04:40 |
| sbalukoff | So, I would have expected the same here. | 04:40 |
| blogan | ah, well this one was probably one of the very early commits in octavia | 04:40 |
| sbalukoff | (Especially scattered throughout our tests.) | 04:40 |
| blogan | if i pushed up a patch that changed the default, would that mess you up a lot? | 04:42 |
| sbalukoff | Probably not-- I will make sure to explicitly set the flag to false in the code I'm doing. | 04:44 |
| blogan | ah, listener controller's put is definitely not calling it either | 04:50 |
| blogan | i mean its not using the flag | 04:51 |
| sbalukoff | Yep. | 04:51 |
| sbalukoff | Did you want to fix that or shall I? | 04:53 |
| blogan | default is changing, it'll get fixed automaticaly | 04:53 |
| blogan | i mean im pushing that change up | 04:53 |
| sbalukoff | Except there's a bug that will happen if you do that without fixing the listener code: In the listener controller right after the to_dict() line we make an assumption about a key that will be in the dict. | 04:54 |
| sbalukoff | What if default_pool_id is not specified. | 04:54 |
| blogan | sbalukoff: ah this may not be as simple, bc the attributes that aren't specified will be Unset type | 04:55 |
| sbalukoff | Yep. | 04:55 |
| sbalukoff | What does that mean "unset type" exactly? | 04:55 |
| blogan | well we'll have to check if that key is in the dict | 04:55 |
| blogan | is a wsme type | 04:55 |
| blogan | literally called Unset | 04:55 |
| sbalukoff | Does the dictionary contain the keys for them that gets rendered? | 04:55 |
| sbalukoff | Ok. | 04:56 |
| blogan | sorry i'm dumb, the to_dict will not put the keys in the returned ict | 04:56 |
| blogan | if they're Unset type | 04:56 |
| sbalukoff | Ok. | 04:56 |
| sbalukoff | So... are you going to fix the listener controller then if you change the default? | 04:57 |
| blogan | yeah, i think thats going to be necessary | 04:58 |
| sbalukoff | Ok, so you're working on that... I'll try and un-fuck the session persistence bug fix. | 04:59 |
| blogan | unless i'm stepping on your toes and causing you more work or causing you to throw away what you're donig | 04:59 |
| sbalukoff | I don't know yet. | 04:59 |
| *** ajmiller has quit IRC | 05:00 | |
| sbalukoff | I don't think so. | 05:00 |
| sbalukoff | So yes: If you have some cycles, please create a patch on master which changes the default of that to_dict method. | 05:00 |
| blogan | okay | 05:00 |
| blogan | this is where our gaps in test coverage scares me | 05:01 |
| sbalukoff | Yeah. | 05:01 |
| sbalukoff | Well, the alternative is to not change the default right now, and just update the put methods. | 05:02 |
| sbalukoff | Maybe look at changing the default after Newton is open? | 05:02 |
| blogan | sbalukoff: so for l7rule, you are checking if values are None, is that because users can specify None or bc you were working around this issue? | 05:02 |
| sbalukoff | After all, any changes to the put methods will be backward compatible since they'll be explicitly setting the parameter. | 05:02 |
| blogan | i can just push the review up and see if anything breaks awfully | 05:02 |
| blogan | i kinda wonder if this is the root cause of many of the out of sync bugs we've seen | 05:03 |
| fnaval | sbalukoff: made more comments on 1547609; it appears that the operating status does to 'ERROR' upon deletion of a member as well. | 05:03 |
| sbalukoff | blogan: Yes, that's what I was working around. | 05:03 |
| fnaval | goes* | 05:03 |
| sbalukoff | fnaval: Saw your previous comment. Have bigger problems right now. | 05:03 |
| blogan | sbalukoff: so you're converting the l7rule to a data model for validation, won't this do the same thing? set the unspecified attributes to None because the data model will init those to None? | 05:07 |
| sbalukoff | blogan: Possibly. Not sure yet. | 05:07 |
| blogan | ok ill step through it | 05:07 |
| sbalukoff | blogan: Are you suggesting not converting to the data model ? | 05:08 |
| blogan | sbalukoff: no, i'm just wondering if you had a quick affirmative/negative to that | 05:09 |
| sbalukoff | blogan: I'm so fried right now I have no idea. | 05:09 |
| blogan | ill step through it, i think since its just validation it won't be an issue | 05:10 |
| *** neelashah has quit IRC | 05:10 | |
| blogan | its fine | 05:19 |
| fnaval | ok, no worries - just adding more info to the bug | 05:22 |
| *** minwang2 has joined #openstack-lbaas | 05:24 | |
| *** minwang2 has quit IRC | 05:26 | |
| blogan | sbalukoff: nvm it might be bad since the data model defaults to invert=False, so that means if invert is not specified... | 05:27 |
| sbalukoff | What does invert do? | 05:27 |
| blogan | ??? | 05:27 |
| blogan | its on l7rule | 05:27 |
| sbalukoff | Oh. | 05:27 |
| sbalukoff | Oh, that's not bad. | 05:28 |
| blogan | if htey're updating a rule that already has invert set to True its not? | 05:28 |
| blogan | or is invert just a one time operation and we dont store it? | 05:28 |
| sbalukoff | We do store it. | 05:28 |
| sbalukoff | If they update a rule and don't specify invert in the update.... | 05:28 |
| sbalukoff | shouldn't it not get changed? | 05:28 |
| blogan | well since its validation, it wont, even though the conversaion to the data model makes it look like invert was passed in as False | 05:29 |
| blogan | but data model is only used for validation, the actual wsme type is used to send to the handler | 05:29 |
| blogan | and validation isn't doing anything with invert, so won't affect it | 05:30 |
| sbalukoff | Right... invert does not matter for that validation. | 05:30 |
| blogan | but a potential problem in the future | 05:30 |
| blogan | not gonna boil the ocean right now though | 05:30 |
| blogan | ah yeah that validaiton is going to have to change though, bc it expects rule type to be there | 05:31 |
| blogan | and for an update, you want to be able to just update the value and nothing else, correct? you have a test for this | 05:32 |
| blogan | althought, perhaps maybe the type should be required on a PUT anyway | 05:32 |
| blogan | no, no | 05:33 |
| blogan | it shouldn't | 05:33 |
| blogan | hmm | 05:33 |
| openstackgerrit | Franklin Naval proposed openstack/octavia: Octavia: Basic LoadBalancer Scenario Test https://review.openstack.org/172199 | 05:33 |
| sbalukoff | blogan: On a put we can flesh out a dummy object with stuff from the database, apply the updates as specified in the PUT command, and validate against that. | 05:36 |
| sbalukoff | I think we already do that, actually. | 05:36 |
| blogan | yeah thats what i'm writing, its not too much work, but probably can be organized abit better | 05:37 |
| blogan | i dont see where its done alreayd though | 05:37 |
| blogan | at least before the validate is called | 05:37 |
| *** fnaval has quit IRC | 05:40 | |
| sbalukoff | Hmmm... Ok. | 05:40 |
| blogan | i put it in the validate call itself, the validate call now takes an optional arg old_l7rule, not sure if you like that but if you dont, say your piece | 05:42 |
| sbalukoff | Er... | 05:43 |
| sbalukoff | I'd rather it didn't. | 05:43 |
| sbalukoff | The validate call should just take a rule object and tell you whether it's valid. | 05:43 |
| sbalukoff | So any prep of that object you need to do, you should do before you try to validate it. | 05:44 |
| sbalukoff | (This way the validation should work cleaner for both creates and updates.) | 05:45 |
| blogan | good point, there is a prepare module that got added for the single create call | 05:46 |
| blogan | might just put it in there, or just leave it in the controller | 05:46 |
| *** kevo has joined #openstack-lbaas | 05:48 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Fix session_persistence deletion bug https://review.openstack.org/287118 | 05:49 |
| *** fnaval has joined #openstack-lbaas | 05:50 | |
| *** armax has joined #openstack-lbaas | 05:50 | |
| sbalukoff | the prepare stuff only applies to creates, where all this stuff should be filled out anyway. | 05:50 |
| sbalukoff | Just leave it in the API controller put method. | 05:50 |
| *** Purandar has quit IRC | 05:52 | |
| blogan | i dont like that, but for the sake of keeping it simple i will | 05:52 |
| *** chenli has joined #openstack-lbaas | 05:57 | |
| *** kevo has quit IRC | 05:57 | |
| *** anilvenkata has joined #openstack-lbaas | 05:58 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Add unit tests for L7 data model additions https://review.openstack.org/290444 | 05:59 |
| sbalukoff | blogan: Are you doing anything to the l7policy controller? | 06:02 |
| *** fnaval has quit IRC | 06:02 | |
| sbalukoff | (ie. am I safe to update it?) | 06:02 |
| *** Purandar has joined #openstack-lbaas | 06:03 | |
| *** fawadkhaliq has joined #openstack-lbaas | 06:07 | |
| *** fnaval has joined #openstack-lbaas | 06:08 | |
| blogan | sbalukoff: yes i believe so far just a one line change | 06:08 |
| sbalukoff | Is it on that validate.sanitize.... call? | 06:09 |
| sbalukoff | If so... then that's not going to work. The assumptions around that sanitize call are all wrong now... :/ | 06:10 |
| sbalukoff | I have to overhaul the whole thing. | 06:11 |
| blogan | sbalukoff: is that only meant to solve for this? | 06:13 |
| sbalukoff | No, it has other purposes as well. | 06:13 |
| sbalukoff | I'm working on a fix for this... you could potentially make your patch dependent on mine when it's ready (later tonight). | 06:14 |
| sbalukoff | So that it passes tox | 06:14 |
| blogan | okay, unless it alreayd passes tox then they can be independent | 06:16 |
| blogan | and it does | 06:16 |
| *** Paco_ has joined #openstack-lbaas | 06:17 | |
| sbalukoff | Ok. | 06:20 |
| openstackgerrit | Brandon Logan proposed openstack/octavia: WIP - Change render_unsets default to False https://review.openstack.org/291013 | 06:23 |
| blogan | lets see how the tempest tests go on it | 06:24 |
| *** fawadkhaliq has quit IRC | 06:30 | |
| *** Purandar has quit IRC | 06:47 | |
| *** evgenyf has joined #openstack-lbaas | 06:54 | |
| *** eranra has joined #openstack-lbaas | 06:55 | |
| *** eranra_ has joined #openstack-lbaas | 06:55 | |
| *** eranra has quit IRC | 06:59 | |
| *** kevo has joined #openstack-lbaas | 07:06 | |
| *** numans has joined #openstack-lbaas | 07:07 | |
| *** prabampm has quit IRC | 07:20 | |
| *** bank has quit IRC | 07:20 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Fix L7Policy update regression https://review.openstack.org/290963 | 07:32 |
| *** rcernin has joined #openstack-lbaas | 07:33 | |
| openstackgerrit | Trevor Vardeman proposed openstack/octavia: Updating unit testing https://review.openstack.org/291041 | 07:51 |
| *** nmagnezi has joined #openstack-lbaas | 07:57 | |
| *** reedip has quit IRC | 08:03 | |
| *** reedip has joined #openstack-lbaas | 08:04 | |
| *** kevo has quit IRC | 08:05 | |
| *** prabampm has joined #openstack-lbaas | 08:07 | |
| *** ihrachys has joined #openstack-lbaas | 08:07 | |
| *** ihrachys has quit IRC | 08:08 | |
| *** [1]evgenyf has joined #openstack-lbaas | 08:13 | |
| *** ihrachys has joined #openstack-lbaas | 08:13 | |
| openstackgerrit | Stephen Balukoff proposed openstack/octavia: Add unit tests for L7 data model additions https://review.openstack.org/290444 | 08:14 |
| *** evgenyf has quit IRC | 08:16 | |
| *** [1]evgenyf is now known as evgenyf | 08:16 | |
| *** armax has quit IRC | 08:17 | |
| *** pcaruana has joined #openstack-lbaas | 08:34 | |
| *** ctracey has quit IRC | 08:36 | |
| *** ctracey has joined #openstack-lbaas | 08:43 | |
| *** kobis has joined #openstack-lbaas | 08:43 | |
| openstackgerrit | Brandon Logan proposed openstack/octavia: Change render_unsets default to False https://review.openstack.org/291013 | 08:49 |
| *** amotoki has quit IRC | 09:15 | |
| *** reedip is now known as reedip_away | 09:15 | |
| *** openstackgerrit has quit IRC | 09:30 | |
| *** chenli has quit IRC | 09:30 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 09:31 | |
| *** openstackgerrit_ is now known as openstackgerrit | 09:31 | |
| *** openstackgerrit has quit IRC | 09:31 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 09:32 | |
| *** openstackgerrit_ is now known as openstackgerrit | 09:32 | |
| *** openstackgerrit has quit IRC | 09:32 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 09:33 | |
| *** openstackgerrit_ is now known as openstackgerrit | 09:33 | |
| *** openstackgerrit has quit IRC | 09:33 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 09:34 | |
| *** openstackgerrit_ is now known as openstackgerrit | 09:34 | |
| *** amotoki has joined #openstack-lbaas | 09:50 | |
| openstackgerrit | Kobi Samoray proposed openstack/octavia: [WIP] Implement custom header support for Octavia https://review.openstack.org/257901 | 10:29 |
| openstackgerrit | Kobi Samoray proposed openstack/octavia: [WIP] Implement custom header support for Octavia https://review.openstack.org/257901 | 10:30 |
| *** yamamoto_ has quit IRC | 10:52 | |
| *** eranra_ has quit IRC | 11:32 | |
| *** yamamoto has joined #openstack-lbaas | 11:44 | |
| *** yamamoto has quit IRC | 11:44 | |
| *** yamamoto has joined #openstack-lbaas | 11:44 | |
| *** yamamoto_ has joined #openstack-lbaas | 11:45 | |
| *** yamamoto has quit IRC | 11:50 | |
| *** evgenyf has quit IRC | 11:56 | |
| *** yamamoto_ has quit IRC | 11:57 | |
| *** yamamoto has joined #openstack-lbaas | 12:28 | |
| *** rtheis has joined #openstack-lbaas | 12:37 | |
| *** doug-fish has quit IRC | 12:38 | |
| *** yamamoto has quit IRC | 12:39 | |
| *** doug-fish has joined #openstack-lbaas | 12:39 | |
| *** yamamoto has joined #openstack-lbaas | 12:43 | |
| *** doug-fish has quit IRC | 12:44 | |
| *** pcaruana has quit IRC | 12:45 | |
| *** evgenyf has joined #openstack-lbaas | 12:46 | |
| *** pcaruana has joined #openstack-lbaas | 13:00 | |
| *** ducttape_ has joined #openstack-lbaas | 13:12 | |
| *** Paco_ has quit IRC | 13:25 | |
| *** doug-fish has joined #openstack-lbaas | 13:26 | |
| *** doug-fish has quit IRC | 13:27 | |
| *** doug-fish has joined #openstack-lbaas | 13:27 | |
| *** evgenyf has quit IRC | 13:27 | |
| *** ducttape_ has quit IRC | 13:29 | |
| *** neelashah has joined #openstack-lbaas | 13:38 | |
| *** links has quit IRC | 13:48 | |
| *** openstackgerrit has quit IRC | 14:03 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 14:04 | |
| *** openstackgerrit_ is now known as openstackgerrit | 14:04 | |
| *** openstackgerrit has quit IRC | 14:04 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 14:05 | |
| *** openstackgerrit_ is now known as openstackgerrit | 14:05 | |
| *** openstackgerrit has quit IRC | 14:05 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 14:06 | |
| *** openstackgerrit_ is now known as openstackgerrit | 14:06 | |
| *** openstackgerrit has quit IRC | 14:07 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 14:07 | |
| *** openstackgerrit_ is now known as openstackgerrit | 14:08 | |
| *** openstackgerrit has quit IRC | 14:08 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 14:08 | |
| *** openstackgerrit_ is now known as openstackgerrit | 14:09 | |
| *** openstackgerrit has quit IRC | 14:09 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 14:10 | |
| *** amotoki has quit IRC | 14:10 | |
| *** openstackgerrit_ is now known as openstackgerrit | 14:10 | |
| *** Guest41345 has joined #openstack-lbaas | 14:11 | |
| *** matt-borland has joined #openstack-lbaas | 14:13 | |
| *** ajmiller has joined #openstack-lbaas | 14:19 | |
| *** nmagnezi has quit IRC | 14:23 | |
| *** nmagnezi has joined #openstack-lbaas | 14:26 | |
| *** pcaruana has quit IRC | 14:41 | |
| *** fnaval has quit IRC | 14:42 | |
| *** ducttape_ has joined #openstack-lbaas | 14:43 | |
| openstackgerrit | Lucas Palm proposed openstack/neutron-lbaas-dashboard: Add the Angular LBaaS V2 'Update Member List' Table Action https://review.openstack.org/290692 | 14:43 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Reload route after performing actions https://review.openstack.org/289686 | 14:47 |
| *** ajmiller has quit IRC | 14:49 | |
| *** pcaruana has joined #openstack-lbaas | 14:53 | |
| openstackgerrit | Lucas Palm proposed openstack/neutron-lbaas-dashboard: Add the Angular LBaaS V2 'Update Member List' Table Action https://review.openstack.org/290692 | 14:59 |
| *** fnaval has joined #openstack-lbaas | 15:05 | |
| *** ajmiller has joined #openstack-lbaas | 15:23 | |
| *** numans has quit IRC | 15:24 | |
| *** numans has joined #openstack-lbaas | 15:26 | |
| *** cody-somerville has joined #openstack-lbaas | 15:41 | |
| *** anilvenkata has quit IRC | 15:45 | |
| *** armax has joined #openstack-lbaas | 16:00 | |
| *** numans has quit IRC | 16:07 | |
| openstackgerrit | Kobi Samoray proposed openstack/octavia: Implement custom header support for Octavia https://review.openstack.org/257901 | 16:19 |
| *** kobis has quit IRC | 16:20 | |
| *** johnsom has quit IRC | 16:27 | |
| *** eranra has joined #openstack-lbaas | 16:34 | |
| *** bank has joined #openstack-lbaas | 16:34 | |
| *** bank has quit IRC | 16:38 | |
| *** ducttape_ has quit IRC | 16:42 | |
| *** matt-borland has quit IRC | 16:45 | |
| *** pcaruana has quit IRC | 16:47 | |
| *** ihrachys has quit IRC | 16:48 | |
| *** matt-borland has joined #openstack-lbaas | 16:50 | |
| *** madhu_ak has joined #openstack-lbaas | 16:53 | |
| *** cody-somerville has quit IRC | 16:55 | |
| *** ducttape_ has joined #openstack-lbaas | 16:55 | |
| *** ducttape_ has quit IRC | 16:59 | |
| *** ihrachys has joined #openstack-lbaas | 17:00 | |
| *** amrith has joined #openstack-lbaas | 17:02 | |
| amrith | rm_work, hello | 17:03 |
| *** Purandar has joined #openstack-lbaas | 17:03 | |
| *** kevo has joined #openstack-lbaas | 17:03 | |
| *** bank has joined #openstack-lbaas | 17:06 | |
| *** Bjoern_ has joined #openstack-lbaas | 17:07 | |
| *** TrevorV has joined #openstack-lbaas | 17:08 | |
| rm_work | hey amrith | 17:09 |
| amrith | hiya | 17:09 |
| rm_work | oh weird, johnsom isn't here | 17:09 |
| rm_work | so technically i am in a meeting right now and really shouldn't be talking here | 17:09 |
| rm_work | but | 17:09 |
| amrith | np, we can catch up later. I'm eating lunch ;) | 17:09 |
| amrith | and mommy always said I should focus on that ... | 17:09 |
| rm_work | ^^ guys, amrith has some experience that may help us with our amp-image building | 17:09 |
| rm_work | heh | 17:09 |
| *** kobis has joined #openstack-lbaas | 17:10 | |
| *** madhu_ak_ has joined #openstack-lbaas | 17:11 | |
| *** matt-borland has quit IRC | 17:12 | |
| *** kobis has quit IRC | 17:16 | |
| openstackgerrit | Merged openstack/octavia: Use consistent type checking of project_id in API https://review.openstack.org/290936 | 17:17 |
| *** alhu has joined #openstack-lbaas | 17:21 | |
| *** prabampm has quit IRC | 17:24 | |
| *** ducttape_ has joined #openstack-lbaas | 17:25 | |
| *** minwang2 has joined #openstack-lbaas | 17:29 | |
| *** fnaval_ has joined #openstack-lbaas | 17:31 | |
| *** fnaval has quit IRC | 17:31 | |
| *** mgarza has joined #openstack-lbaas | 17:40 | |
| openstackgerrit | Merged openstack/octavia: Change render_unsets default to False https://review.openstack.org/291013 | 17:42 |
| *** kobis has joined #openstack-lbaas | 17:42 | |
| *** kobis has quit IRC | 17:42 | |
| *** johnsom has joined #openstack-lbaas | 17:45 | |
| *** fawadkhaliq has joined #openstack-lbaas | 17:47 | |
| *** SumitNaiksatam has joined #openstack-lbaas | 17:49 | |
| *** yamamoto has quit IRC | 17:49 | |
| *** johnsom has quit IRC | 17:51 | |
| *** johnsom has joined #openstack-lbaas | 17:52 | |
| *** johnsom has quit IRC | 17:58 | |
| *** johnsom has joined #openstack-lbaas | 17:58 | |
| *** johnsom has quit IRC | 18:02 | |
| *** johnsom has joined #openstack-lbaas | 18:02 | |
| *** kevo has quit IRC | 18:05 | |
| sbalukoff | Nice! | 18:06 |
| sbalukoff | amrith: Have you done much with minimal images (like ubuntu-core or whatever it is)? | 18:06 |
| fnaval_ | sbalukoff: thanks for your comment on 1547609 - I re-clarified the bug with another comment. | 18:07 |
| fnaval_ | also thanks for the review on the lb scenario test | 18:07 |
| johnsom | rm_work Sorry I missed you, patching reboot action | 18:08 |
| fnaval_ | there were some reason why .testr.conf was made to point to the unit directory; I'll comment that in the review | 18:08 |
| *** fnaval_ is now known as fnaval | 18:09 | |
| *** numans has joined #openstack-lbaas | 18:09 | |
| sbalukoff | fnaval: I'm concerned that that change eliminates the functional tests from the py27 and py34 environment runs. | 18:10 |
| sbalukoff | That is perhaps fix-able with a further update to the tox.ini | 18:10 |
| *** bank has quit IRC | 18:10 | |
| sbalukoff | But right now, you can even see in the jenkins tests that were run against the code tree that there are hundreds of functional tests that didn't get run. | 18:11 |
| *** nmagnezi has quit IRC | 18:12 | |
| fnaval | sbalukoff: yea... the problem was that CONF.debug in Octavai was getting overridden by Tempest thus causing some octavia unit tests to break | 18:13 |
| rm_work | johnsom: amrith had ideas on fixing our image building on every run thing.... better than the CDN pre-built option but still way faster | 18:13 |
| fnaval | there must be a hacky way to get around it; i'll have to dig through testr docs or other projects implementations tos ee how to get around it | 18:13 |
| rm_work | johnsom: i was going to point him to you | 18:14 |
| johnsom | Ok, cool. Yeah, interested to hear. | 18:14 |
| *** alhu has quit IRC | 18:15 | |
| *** jwarendt has joined #openstack-lbaas | 18:16 | |
| sbalukoff | fnaval: Thanks. Also, I've responded to your comments on that bug report. | 18:19 |
| *** ihrachys has quit IRC | 18:23 | |
| *** anilvenkata has joined #openstack-lbaas | 18:23 | |
| *** ducttape_ has quit IRC | 18:26 | |
| *** ducttape_ has joined #openstack-lbaas | 18:26 | |
| *** openstackgerrit_ has joined #openstack-lbaas | 18:30 | |
| sbalukoff | Hey folks-- it looks like I'm going to be unavailable most of today as I deal with a bunch of internal noise. However, if anyone has some cycles, I would appreciate some eyes on this, as I think it's also a fairly critical bugfix: https://review.openstack.org/#/c/290963/ | 18:33 |
| fnaval | anyone know what "NO_MONITOR" operating status means in the context of Octavia? | 18:34 |
| sbalukoff | fnaval: Yes, it means that no health monitor is active for that member. | 18:35 |
| sbalukoff | I think that's what is supposed to happen when there is no health monitor on the pool. | 18:35 |
| fnaval | hmm.. sbalukoff so that's what im seeing for both members | 18:37 |
| fnaval | but i'm only shutting down one of them | 18:38 |
| fnaval | both members exist in the same pool | 18:38 |
| sbalukoff | fnaval: Is there a health monitor on the pool? | 18:38 |
| fnaval | but 1 is disabled | 18:38 |
| fnaval | yes a health monitor is one the pool | 18:38 |
| sbalukoff | Ok, that's screwey. | 18:38 |
| fnaval | is NO_MONITOR a pending state? | 18:38 |
| sbalukoff | I'm not sure. | 18:38 |
| fnaval | ok, so maybe the bug now is that shutting down a member in a pool causing all members to go into NO_MONITOR | 18:39 |
| *** matt-borland has joined #openstack-lbaas | 18:39 | |
| sbalukoff | fnaval: That would be a bug. | 18:39 |
| fnaval | ok - i'm going to add more debugging also; i'll change the title of the bug then to reflect it? or should i file new bug? | 18:41 |
| sbalukoff | No, change the title of the existing bug. | 18:42 |
| sbalukoff | Unless you want to get rid of the history on the current bug. | 18:42 |
| *** alhu has joined #openstack-lbaas | 18:43 | |
| fnaval | yea that might be better to file new plus the old bug might be valid later | 18:44 |
| *** SumitNaiksatam has left #openstack-lbaas | 18:46 | |
| *** yamamoto has joined #openstack-lbaas | 18:50 | |
| rm_work | err, amrith here | 18:50 |
| rm_work | johnsom: amrith amrith: johnsom | 18:51 |
| amrith | johnsom, hello | 18:51 |
| rm_work | bbl meeting | 18:51 |
| amrith | rm_work, yes. saw the flashing icons and noticed -infra first. | 18:51 |
| rm_work | heh | 18:51 |
| amrith | johnsom, now a good time? | 18:51 |
| johnsom | amrith Yes | 18:51 |
| amrith | okey dokey | 18:52 |
| amrith | sbalukoff, just saw your question as well. | 18:52 |
| amrith | so, if I understand correctly, rm_work was asking about trove and tarballs/images that we build | 18:52 |
| *** TrevorV has quit IRC | 18:52 | |
| amrith | are you familiar with trove, or do you want the 2c tour? | 18:52 |
| johnsom | I know Trove is dbaas with various database images. That is about it | 18:53 |
| *** bank has joined #openstack-lbaas | 18:53 | |
| amrith | yes, that's a good start | 18:53 |
| amrith | common API for all things you can do (database agnostic) | 18:53 |
| amrith | database guest images contain an OS, a database of your choice and a guest agent | 18:53 |
| amrith | that guest agent is database specific | 18:53 |
| amrith | in production, one assumes that you would have images that have the OS+database+agent baked on. | 18:54 |
| amrith | but in development you have different desires | 18:54 |
| amrith | so what we do is that sample elements (we use dib) will generate an image for a database (say MySQL 5.6) | 18:54 |
| amrith | and as a developer, I'm iterating on the controller and guest agent code to do something with MySQL 5.6 | 18:55 |
| amrith | say replication. | 18:55 |
| amrith | so I make a change that straddles controller and guest. | 18:55 |
| amrith | it is a bore if I have to rebuild the image for that | 18:55 |
| amrith | so instead, the guest image (for development) has an upstart job that rsyncs the guest code at launch | 18:55 |
| amrith | that way, development cycles are very fast. | 18:55 |
| amrith | and we publish those guests (development use only) onto tarballs.o.o | 18:55 |
| *** ducttape_ has quit IRC | 18:56 | |
| amrith | am I maybe in the right general zipcode as the questions you have? | 18:56 |
| *** Aish has joined #openstack-lbaas | 18:56 | |
| johnsom | I am following you. I didn't have a specific question. We use DIB as well. | 18:57 |
| johnsom | Interesting on the rsync. That rsyncs back to the devstack host? | 18:58 |
| amrith | yes, let me find you the element. one second | 18:58 |
| johnsom | I think the concern folks have here is our scenario gate builds the image (of course to test the agent code) which takes ~5 minutes | 18:59 |
| *** yamamoto has quit IRC | 18:59 | |
| amrith | oh, don't be concerned about that ;) our gate builds the image each time and it takes 20'ish minutes ... | 19:00 |
| *** kevo has joined #openstack-lbaas | 19:02 | |
| amrith | ok, found it. | 19:02 |
| amrith | the rsync (for ubuntu guests) comes from https://github.com/openstack/trove-integration/blob/master/scripts/files/trove-guest.upstart.conf | 19:02 |
| amrith | https://github.com/openstack/trove-integration/blob/master/scripts/files/trove-guest.upstart.conf#L19 | 19:03 |
| johnsom | I will take a look at that | 19:03 |
| amrith | as far as using images that the gate could use, we went that route for a while | 19:03 |
| amrith | there were some issues with that | 19:03 |
| amrith | consider this, the guest image for gate/check could use the same 'rsync the guest code and run tests' model. | 19:04 |
| johnsom | Yeah, I am pretty much against that. | 19:04 |
| amrith | however, the issue we ran into was that as projects (python projects outside openstack) or other openstack projects were updated, the requirements on the guest were different | 19:04 |
| johnsom | Canned images are to likely to become a problem | 19:04 |
| amrith | and since we want to remain consistent with g-r, we were ending up having to update that 'cached' image often | 19:04 |
| amrith | and when we didn't we had hard to diagnose test failures. | 19:05 |
| johnsom | Right | 19:05 |
| amrith | so while it meant a cost of 15'ish minutes per gate job, we build a guest image on the fly. | 19:05 |
| amrith | I don't know a good way to get both. i.e. benefits of caching without the cost. | 19:05 |
| amrith | put differently, no such thing as a free lunch | 19:05 |
| johnsom | Which base are you using for Ubunut? | 19:07 |
| * amrith looks | 19:08 | |
| johnsom | We are using the base "ubuntu" element. About four-five months ago I tried minimal and core, but it seemed like both were broken | 19:08 |
| openstackgerrit | Merged openstack/neutron-lbaas: Make all tox targets constrained https://review.openstack.org/288431 | 19:09 |
| amrith | cd .. | 19:09 |
| *** numans has quit IRC | 19:09 | |
| *** madhu_ak_ has quit IRC | 19:09 | |
| amrith | we use ubuntu | 19:11 |
| amrith | https://github.com/openstack/trove-integration/blob/master/scripts/functions_qemu#L54 | 19:11 |
| amrith | $DISTRO is 'ubuntu' | 19:11 |
| sbalukoff | Dang. I was hoping someone else had gotten smaller images working. | 19:11 |
| amrith | or 'redhat' | 19:11 |
| amrith | well, what we do is this | 19:12 |
| johnsom | Yep, so same as what we do | 19:12 |
| amrith | we install ubuntu and then trim packages | 19:12 |
| amrith | https://github.com/openstack/trove-integration/blob/master/scripts/files/elements/ubuntu-guest/pre-install.d/01-trim-pkgs | 19:12 |
| johnsom | Yeah, I have considered doing that. I just wasn't sure how much we would save vs. the headache of dealing with the dependency issues | 19:14 |
| *** ducttape_ has joined #openstack-lbaas | 19:17 | |
| *** fnaval has quit IRC | 19:21 | |
| *** fnaval has joined #openstack-lbaas | 19:22 | |
| *** fnaval has quit IRC | 19:22 | |
| *** bana_k has joined #openstack-lbaas | 19:22 | |
| *** bank has left #openstack-lbaas | 19:23 | |
| *** bana_k has left #openstack-lbaas | 19:23 | |
| *** bank has joined #openstack-lbaas | 19:23 | |
| *** bank has quit IRC | 19:23 | |
| *** bank has joined #openstack-lbaas | 19:23 | |
| *** fnaval has joined #openstack-lbaas | 19:24 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/neutron-lbaas: Updated from global requirements https://review.openstack.org/291376 | 19:27 |
| *** kobis has joined #openstack-lbaas | 19:40 | |
| *** rcernin has quit IRC | 19:40 | |
| openstackgerrit | Justin Pomeroy proposed openstack/neutron-lbaas-dashboard: Hide load balancers v2 panel if not enabled https://review.openstack.org/291388 | 19:41 |
| *** armax has quit IRC | 19:46 | |
| *** SumitNaiksatam has joined #openstack-lbaas | 19:50 | |
| openstackgerrit | Justin Pomeroy proposed openstack/neutron-lbaas-dashboard: Hide load balancers v2 panel if not enabled https://review.openstack.org/291388 | 19:59 |
| *** kobis has quit IRC | 20:00 | |
| *** kobis has joined #openstack-lbaas | 20:01 | |
| openstackgerrit | Justin Pomeroy proposed openstack/neutron-lbaas-dashboard: Hide load balancers v2 panel if not enabled https://review.openstack.org/291388 | 20:02 |
| *** kobis has quit IRC | 20:08 | |
| *** alhu has quit IRC | 20:09 | |
| *** alhu has joined #openstack-lbaas | 20:10 | |
| *** TrevorV|Home has joined #openstack-lbaas | 20:13 | |
| *** diogogmt has joined #openstack-lbaas | 20:13 | |
| *** TrevorV has joined #openstack-lbaas | 20:17 | |
| *** ducttape_ has quit IRC | 20:18 | |
| *** TrevorV|Home has quit IRC | 20:19 | |
| *** TrevorV|Home has joined #openstack-lbaas | 20:19 | |
| *** armax has joined #openstack-lbaas | 20:20 | |
| TrevorV|Home | blogan, rm_work is rax irc down> | 20:20 |
| TrevorV|Home | down?*** | 20:20 |
| *** ducttape_ has joined #openstack-lbaas | 20:20 | |
| xgerman | mmh | 20:21 |
| xgerman | maybe we need to let you on our HP hip chat? | 20:22 |
| TrevorV|Home | No way, you guys will start adding tasks and stuff, and then I'll slip through the cracks and end up working for HP without a paycheck | 20:22 |
| TrevorV|Home | Not happenin broski | 20:22 |
| johnsom | I like that idea | 20:23 |
| TrevorV|Home | Well, its not on the table | 20:23 |
| TrevorV|Home | At a minimum "no paycheck" isn't on the table. | 20:24 |
| johnsom | But we won't shun you... We will talk with you.... Grin | 20:24 |
| TrevorV|Home | I feel like that goes with a chant somewhere... "One of us... One of us" | 20:24 |
| TrevorV|Home | :P | 20:24 |
| TrevorV|Home | btw johnsom did you see my review from last night? The unit test? | 20:25 |
| johnsom | I did. I am working down the list as we type | 20:25 |
| TrevorV|Home | Cool. | 20:25 |
| johnsom | Thanks! | 20:25 |
| TrevorV|Home | No problem | 20:25 |
| TrevorV|Home | I did miss one test (on purpose) because I feel like the test is worthless. | 20:26 |
| TrevorV|Home | The "coverage" point you made concerning the try/catch around the handler accepting the request. | 20:26 |
| TrevorV|Home | The test would be "pass in valid object, but handler has an error", and mocking handler call to "fail" seems like I'm engineering the test to pass, not to simulate a test scenario. | 20:26 |
| johnsom | Hmm, there are code changes in here that aren't in the commit message | 20:27 |
| TrevorV|Home | Oh, my bad, I'll update that. | 20:27 |
| TrevorV|Home | I've got a few more things to add to the doc update too and that'll go up in a review | 20:27 |
| *** TrevorV has quit IRC | 20:28 | |
| openstackgerrit | Merged openstack/octavia: Use rfc3986 module for URL validations https://review.openstack.org/290918 | 20:28 |
| openstackgerrit | Lucas Palm proposed openstack/neutron-lbaas-dashboard: Do not allow deleting a listener that has a default pool https://review.openstack.org/291415 | 20:31 |
| *** anilvenkata has quit IRC | 20:33 | |
| johnsom | TrevorV|Home Ok, just the health_monitor create is missing, otherwise looks ok | 20:36 |
| openstackgerrit | Merged openstack/neutron-lbaas-dashboard: Fix gate scripts to use script dir for sourcing commons https://review.openstack.org/290037 | 20:37 |
| TrevorV|Home | johnsom, I'm sorry, what? | 20:39 |
| TrevorV|Home | We don't have a repository test to create a HM> | 20:39 |
| TrevorV|Home | HM?*** | 20:39 |
| johnsom | http://sonar.octavia.io/archive/octavia/85fea8185b014e68d7cdcbcd610f7346af9d438c/663/cover/octavia_db_repositories_py.html#n291 | 20:41 |
| TrevorV|Home | Oooooh that's a diff spot than I remember... Its a healthmonitor on the redirect pool. | 20:42 |
| TrevorV|Home | I'll add that to the test | 20:42 |
| TrevorV|Home | I made sure to throw the other HM on there, my bad. I'll fix it | 20:43 |
| TrevorV|Home | And the mention of the changes in the actual method | 20:43 |
| TrevorV|Home | my b | 20:43 |
| johnsom | NP | 20:43 |
| rm_work | sbalukoff: NICE i didn't see that rfc3986 thing come in, but that's excellent -- i have been trying to get us out of the string processing business >_< | 20:44 |
| *** sbalukoff has quit IRC | 20:45 | |
| openstackgerrit | Lucas Palm proposed openstack/neutron-lbaas-dashboard: Do not allow deleting a listener that has a default pool https://review.openstack.org/291415 | 20:54 |
| *** amrith is now known as _amrith_ | 21:01 | |
| TrevorV|Home | no balukoff | 21:07 |
| TrevorV|Home | dammit | 21:07 |
| *** matt-borland has quit IRC | 21:15 | |
| openstackgerrit | Trevor Vardeman proposed openstack/octavia: Single-Create Documentation https://review.openstack.org/291442 | 21:21 |
| TrevorV|Home | nvmd I figured it out | 21:21 |
| TrevorV|Home | Also, johnsom ^^ I'm sure its not as much description as we want, but it made sense | 21:22 |
| TrevorV|Home | to me at least | 21:22 |
| johnsom | TrevorV Sorry, the last few weeks have been a blur, are we using the same endpoint for get-an-lb? ``URL/v1/loadbalancers`` | 21:25 |
| johnsom | TrevorV|Home | 21:28 |
| *** neelashah has quit IRC | 21:28 | |
| TrevorV|Home | yes | 21:29 |
| TrevorV|Home | johnsom, | 21:29 |
| johnsom | Sorry, the last few weeks have been a blur, are we using the same endpoint for get-an-lb? ``URL/v1/loadbalancers`` | 21:29 |
| TrevorV|Home | yes johnsom | 21:29 |
| johnsom | Ok, then I think we can just change the section divider and this should be fine. | 21:30 |
| openstackgerrit | Trevor Vardeman proposed openstack/octavia: Updating unit testing https://review.openstack.org/291041 | 21:33 |
| TrevorV|Home | johnsom, that was the section divider used for different calls all over the place, but I'm not against changing it. | 21:38 |
| TrevorV|Home | Is +++++++ going to provide anything weird though? | 21:38 |
| johnsom | Well, with the same divider it looks like it is missing the endpoint definition, i.e. a new endpoint. I was thinking maybe we could make it more clear that this is with the same endpoint | 21:39 |
| johnsom | I have no idea what it will look like with ++++, It's just the next level down. Just a minute, I can try it out | 21:39 |
| TrevorV|Home | okay | 21:41 |
| johnsom | TrevorV|Home The +++'s renders as just a slightly smaller font. I think the +++'s is a good idea | 21:44 |
| TrevorV|Home | Alrighty, I'll throw that in there then. | 21:44 |
| openstackgerrit | Trevor Vardeman proposed openstack/octavia: Single-Create Documentation https://review.openstack.org/291442 | 21:45 |
| TrevorV|Home | Heh, used the website to do the edit... GG | 21:45 |
| johnsom | Hey, it works | 21:46 |
| TrevorV|Home | It worked great, for something like that. | 21:47 |
| *** ihrachys has joined #openstack-lbaas | 21:49 | |
| johnsom | dougwig FYI, I could still use a second set of eyes on this: https://review.openstack.org/#/c/288797/ | 21:55 |
| *** neelashah has joined #openstack-lbaas | 21:56 | |
| dougwig | johnsom: you never responded to my comment about it, so i was waiting. also, you seem to have some failures now. :) | 21:58 |
| johnsom | Yeah, it is the rootwrap issue I can't figure out. It doesn't reproduce local for me | 21:59 |
| *** doug-fish has quit IRC | 21:59 | |
| *** jschwarz has joined #openstack-lbaas | 21:59 | |
| johnsom | Wondering if you have any ideas | 22:00 |
| *** alhu has quit IRC | 22:10 | |
| *** rcernin has joined #openstack-lbaas | 22:11 | |
| *** alhu has joined #openstack-lbaas | 22:12 | |
| dougwig | johnsom: sigh, that is not at all what i meant. i meant, why isn't the payload for *) also acceptable for what you're trying to do? | 22:12 |
| *** jschwarz has quit IRC | 22:12 | |
| dougwig | :) | 22:12 |
| johnsom | Ok | 22:12 |
| *** _amrith_ is now known as amrith | 22:14 | |
| dougwig | johnsom: added another comment. if you reply with another bash syntax tutorial, i will snort dr. pepper in humor. | 22:19 |
| johnsom | Challenge accepted | 22:19 |
| *** rcernin has quit IRC | 22:22 | |
| *** alhu has quit IRC | 22:26 | |
| *** woodster_ has joined #openstack-lbaas | 22:27 | |
| johnsom | dougwig You do realize you aren't actually commenting on the part I'm having trouble with right? | 22:27 |
| johnsom | Which is probably by design... | 22:28 |
| johnsom | http://logs.openstack.org/97/288797/10/check/gate-neutron-lbaasv2-dsvm-api-namespace-nv/550ae5a/logs/screen-q-lbaasv2.txt.gz#_2016-03-08_17_32_41_138 | 22:28 |
| dougwig | yes. rootwrap is a mystery to me. | 22:28 |
| johnsom | But the right filters are present in the test archive, works local ok | 22:28 |
| johnsom | Joy. Ok, so apply head to wall, repeat. | 22:29 |
| dougwig | johnsom: 3:29 PM <kevinbenton> yeah, that means its making a call its not allwoed to | 22:30 |
| dougwig | did we prune out the rootwrap filters or something? | 22:30 |
| dougwig | johnsom: is that setup actually installing the rootwrap file from neutron-lbaas ? | 22:31 |
| johnsom | No, I double checked. The Line above that link is the call "ip netns exec <> haproxy which is valid here: http://logs.openstack.org/97/288797/10/check/gate-neutron-lbaasv2-dsvm-api-namespace-nv/550ae5a/logs/etc/neutron/rootwrap.d/lbaas-haproxy.filters.txt.gz | 22:31 |
| johnsom | As far as I can see it hasn't been touched in a long time | 22:31 |
| dougwig | you could always take kevin's tongue-in-cheek advice: 3:29 PM <kevinbenton> the answer is to change your config to 'root_helper = sudo' | 22:31 |
| *** diogogmt has quit IRC | 22:32 | |
| johnsom | Yes, temping. I wonder if that is getting set different local for me vs. the gate. I will double check that. | 22:33 |
| kevinbenton | johnsom: do you have a locally compiled haproxy? | 22:36 |
| kevinbenton | johnsom: i have run into issues in the past when i was changing dnsmasq where the filter would not match | 22:36 |
| johnsom | No | 22:36 |
| dougwig | kevinbenton: no, should be an ubuntu package. | 22:36 |
| johnsom | It's upstream. | 22:36 |
| kevinbenton | johnsom: what does 'which haproxy' return? | 22:37 |
| johnsom | kevinbenton Is there something you can think of why this work run fine on a local devstack install but not on the gate? | 22:37 |
| johnsom | /usr/sbin/haproxy | 22:38 |
| kevinbenton | johnsom: for some reason in the gate haproxy may not be in the allowed exec dirs | 22:38 |
| kevinbenton | http://logs.openstack.org/97/288797/10/check/gate-neutron-lbaasv2-dsvm-api-namespace-nv/550ae5a/logs/etc/neutron/rootwrap.conf.txt.gz | 22:38 |
| *** chlong has quit IRC | 22:39 | |
| dougwig | if we are completely out of debugging ideas, we can ask infra to hold a dsvm test node for us to examine. | 22:39 |
| dougwig | johnsom: ^^ | 22:39 |
| johnsom | Ok, I can do a hack to check that | 22:40 |
| dougwig | the bullets in that particular gun reload slowly, so it should be a last resort. | 22:40 |
| johnsom | Ok. I can do some logging hacks before we need to use up a bullet | 22:40 |
| dougwig | i wonder if haproxy revved at some point in the ubuntu that's running the test, too. | 22:42 |
| kevinbenton | dougwig, johnsom: might be a stupid question, but is haproxy installed on that? | 22:43 |
| kevinbenton | dougwig: i don't see a mention of the install in the console | 22:43 |
| *** rtheis has quit IRC | 22:44 | |
| *** Bjoern_ has quit IRC | 22:44 | |
| kevinbenton | dougwig, johnsom: http://logs.openstack.org/97/288797/10/check/gate-neutron-lbaasv2-dsvm-api-namespace-nv/550ae5a/logs/devstacklog.txt.gz#_2016-03-08_17_17_09_887 | 22:44 |
| kevinbenton | no mention of haproxy in those | 22:44 |
| johnsom | Wow, What the heck??? It's not in the dpkgs list either | 22:45 |
| johnsom | Well, that is interesting | 22:45 |
| johnsom | kevinbenton Back story here, I'm trying to breathe life into a new gate to test old code... | 22:46 |
| kevinbenton | johnsom: makes sense | 22:46 |
| kevinbenton | johnsom: sounds like you need a devstack flag of some sort to add haproxy to the install list | 22:47 |
| kevinbenton | rootwrap could also use an improvement to say that it couldn't find the command at all :/ | 22:47 |
| johnsom | Yeah, I will have to figure out how that is missing here. Thanks for the second set of eyes! | 22:47 |
| dougwig | how did the loadbalancer tests pass without it?!?!? holy heck. | 22:47 |
| johnsom | Well, it doesn't *actually* touch haproxy until there is a listener | 22:48 |
| johnsom | It's all DB games | 22:48 |
| TrevorV|Home | xgerman, johnsom ExtensionsNotFound: Extensions not found: ['n-lbaasv2-cascade-delete']. | 22:48 |
| TrevorV|Home | How do I fix that? | 22:48 |
| TrevorV|Home | Config file entry? | 22:49 |
| dougwig | oh, hahahahahahahahaha. | 22:49 |
| xgerman | it got pulled from the release | 22:49 |
| dougwig | oh, bitter burn. | 22:49 |
| TrevorV|Home | Okay, well devstack be broked then | 22:50 |
| TrevorV|Home | Or something. | 22:50 |
| TrevorV|Home | Maybe | 22:50 |
| TrevorV|Home | Using my patch in devstack is broke at the least. | 22:50 |
| TrevorV|Home | my neutron-lbaas one | 22:50 |
| dougwig | how do the v1 tests pass? | 22:50 |
| TrevorV|Home | https://review.openstack.org/#/c/257201/ | 22:50 |
| TrevorV|Home | that one | 22:50 |
| dougwig | johnsom: v1? | 22:51 |
| johnsom | dougwig I had the same question. I will dig into that | 22:51 |
| *** alhu has joined #openstack-lbaas | 22:51 | |
| *** alhu has quit IRC | 22:54 | |
| *** mgarza has quit IRC | 22:54 | |
| johnsom | dougwig It looks like the v1 tests suck | 22:55 |
| johnsom | http://logs.openstack.org/97/288797/10/check/gate-neutron-lbaasv1-dsvm-api/2fc349a/logs/screen-q-lbaas.txt.gz | 22:55 |
| kevinbenton | johnsom: let me know if that does turn out to be the issue. oslo_rootwrap is supposed to have a separate exception for this | 22:56 |
| kevinbenton | johnsom: FilterMatchNotExecutable | 22:56 |
| johnsom | kevinbenton Ok, will do | 22:56 |
| kevinbenton | johnsom: so either it is installed and there is another issue | 22:56 |
| kevinbenton | johnsom: or oslo rootwrap has a bug and is returning the wrong exception | 22:56 |
| *** doug-fish has joined #openstack-lbaas | 22:59 | |
| dougwig | johnsom: haproxy install should be in the devstack plugin log. one sec. | 23:02 |
| *** doug-fish has quit IRC | 23:03 | |
| dougwig | johnsom: and it's not there. | 23:03 |
| dougwig | johnsom: v1 uses internal devstack lbaas code, v2 uses our plugin. i think it's a devstack plugin bug. | 23:03 |
| *** doug-fish has joined #openstack-lbaas | 23:03 | |
| *** diogogmt has joined #openstack-lbaas | 23:04 | |
| johnsom | Yep | 23:05 |
| johnsom | the install function never gets called | 23:05 |
| neelashah | ajmiller - hi | 23:07 |
| *** doug-fish has quit IRC | 23:08 | |
| ajmiller | neelashah hi | 23:08 |
| neelashah | ajmiller: doug-fish indicated you were planning to review the last two patches for the lbaas dashboard? we need to close the work today for the RC1 tomorrow | 23:08 |
| ajmiller | Yes, I am looking at them right now. | 23:08 |
| neelashah | ajmiller - super, thanks | 23:09 |
| ajmiller | So far everything looks good with them, but I have a few more cases to test. | 23:09 |
| *** alhu has joined #openstack-lbaas | 23:09 | |
| openstackgerrit | Michael Johnson proposed openstack/neutron-lbaas: Fix the neutron-lbaas namespace driver gate https://review.openstack.org/288797 | 23:11 |
| *** johnsom_ has joined #openstack-lbaas | 23:12 | |
| *** johnsom__ has joined #openstack-lbaas | 23:13 | |
| *** doug-fish has joined #openstack-lbaas | 23:14 | |
| *** ducttape_ has quit IRC | 23:15 | |
| neelashah | ajmiller - np, I just wanted to check where it was on your list…I need to let johnsom know once these two patches merge that we are ready to release | 23:15 |
| *** johnsom has quit IRC | 23:16 | |
| *** ihrachys has quit IRC | 23:16 | |
| *** johnsom_ has quit IRC | 23:17 | |
| ajmiller | OK. | 23:18 |
| *** doug-fish has quit IRC | 23:19 | |
| *** neelashah has quit IRC | 23:19 | |
| *** alhu has quit IRC | 23:23 | |
| *** alhu has joined #openstack-lbaas | 23:25 | |
| *** alhu has quit IRC | 23:26 | |
| bank | Hi team , this one is valid bug right ? https://bugs.launchpad.net/octavia/+bug/1555805 | 23:36 |
| openstack | Launchpad bug 1555805 in octavia "octavia get loadbalancer api does not return the load balancer list" [Undecided,New] | 23:36 |
| openstackgerrit | Trevor Vardeman proposed openstack/octavia: Single-Create Documentation https://review.openstack.org/291442 | 23:37 |
| *** alhu has joined #openstack-lbaas | 23:37 | |
| johnsom__ | bank No, you are curling the v1 API but creating a v2 load balancer | 23:38 |
| bank | oops ! | 23:38 |
| bank | got it . thanks. | 23:38 |
| fnaval | speaking of bugs, bug 1555854 | 23:39 |
| openstack | bug 1555854 in octavia "Members go into ERROR operating status upon adding Health Monitor" [Undecided,New] https://launchpad.net/bugs/1555854 | 23:39 |
| bank | johnsom__ http://octavia.io/review/master/api/octaviaapi.html | 23:40 |
| bank | I am taking about the octavia api | 23:40 |
| johnsom__ | Oh, you are going direct to octavia API, not using neutron API | 23:40 |
| bank | yes | 23:40 |
| fnaval | related to the health monitor delete issue i was seeing in recent days. it's actually an issue with how members statuses change on the addition of a health monitor; confirmed code issue | 23:40 |
| bank | http://octavia.io/review/master/api/octaviaapi.html#list-load-balancers | 23:41 |
| johnsom__ | bank It could be an issue with the tenant being different. | 23:47 |
| bank | yes when I try to get the LBs projectid/tenant is empty | 23:48 |
| bank | so its not able to pull the data | 23:48 |
| johnsom__ | I think you can put a parameter on the curl call with the tenant id | 23:49 |
| bank | I ll try that out. | 23:49 |
| johnsom__ | I still thought that going direct to octavia without a tenant/project would return all of them | 23:50 |
| *** johnsom__ has quit IRC | 23:51 | |
| *** johnsom has joined #openstack-lbaas | 23:51 | |
| johnsom | blogan might have better info | 23:51 |
| bank | oh ok, as of now Its not returning load_balancers = self.repositories.load_balancer.get_all( | 23:52 |
| bank | context.session, project_id=project_id) | 23:52 |
| *** alhu has quit IRC | 23:52 | |
| bank | it applies that project id filter in get_all of repositories.py file so we are not getting anything , | 23:56 |
| openstackgerrit | Michael Johnson proposed openstack/neutron-lbaas: Fix the neutron-lbaas namespace driver gate https://review.openstack.org/288797 | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!