| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2017-01-18 | |||
| 08:04:32 | RuiChen | default quota need project id too, it's passed by project=res.id in your other patch | |
| 08:05:36 | reedip | sorry , wrong statement. It needs project id but not in the URI | |
| 08:05:51 | reedip | thats why i put requires_id as false for default quotas in the second patch | |
| 08:06:07 | reedip | I was typing faster than I was thinking, thats why the wrong statement :D | |
| 08:10:43 | RuiChen | never mind, I debug the patch in my devstack to try openstack quota show, like this | |
| 08:10:50 | RuiChen | https://www.irccloud.com/pastebin/LaHmzLVV/ | |
| 08:12:59 | reedip | this is quota show, right ? | |
| 08:13:07 | RuiChen | yes | |
| 08:13:08 | reedip | try quota set and quota-show --default | |
| 08:13:36 | reedip | this is your code ? | |
| 08:14:19 | RuiChen | https://www.irccloud.com/pastebin/Lpl0Srf1/ | |
| 08:14:29 | RuiChen | this quota set --network -1 xxx | |
| 08:14:55 | RuiChen | quota show --default have to apply your other patch | |
| 08:15:45 | reedip | RuiChen, how did the request body have ID when you deleted it using the if loop ? | |
| 08:17:53 | RuiChen | yeah, have to handle the resource_key in body, then delete 'id' | |
| 08:19:01 | reedip | RuiChen : and you are trying this with SDK 0.9.12 right ??? :) | |
| 08:19:11 | reedip | just confirming , please do not mind :) | |
| 08:20:19 | reedip | Actually with 0.9.12, https://www.irccloud.com/pastebin/Lpl0Srf1/ was not happening for me, without my patch. It was failing at the _prepare_requests because quotas didnt have the ID | |
| 08:20:49 | RuiChen | I try the patch in current openstack sdk master, I think the code same with 0.9.12 | |
| 08:21:14 | reedip | :) | |
| 08:21:14 | reedip | Yes, it should be | |
| 08:24:06 | RuiChen | if we can call the super class _prepare_request(), then delete 'id' in body or body[resource_key] to handle the special case for quota and default quota, i think it's better solution | |
| 08:24:46 | RuiChen | you don't have to care the _prepare_request() change in the future. | |
| 08:25:02 | reedip | RuiChen : wait, I am trying your change | |
| 08:25:17 | reedip | I am cloning the latest release and putting your change | |
| 08:25:28 | reedip | I still feel I will get a bug :P | |
| 08:28:53 | RuiChen | : ), I'm trying to make small change to handle quota and default quota (show and set action) | |
| 08:31:42 | reedip | RuiChen : I think there has been some change somewhere :D | |
| 08:35:55 | reedip | RuiChen: found the bug ! | |
| 08:35:55 | RuiChen | finally, it worked | |
| 08:35:59 | reedip | :) | |
| 08:36:03 | reedip | Wait RuiChen :) | |
| 08:36:31 | reedip | I had 0.9.10 :P | |
| 08:36:47 | reedip | thats why the above code worked. Now using 0.9.12 and ur patch ! | |
| 08:36:54 | reedip | and got the issue | |
| 08:37:09 | RuiChen | https://www.irccloud.com/pastebin/xFXl6TFa/ | |
| 08:37:09 | RuiChen | https://www.irccloud.com/pastebin/H5yOdoYA/ | |
| 08:39:49 | reedip | Ok, now it worked ! | |
| 08:40:33 | RuiChen | \o/ | |
| 08:41:14 | RuiChen | thank you mention the quota set case, that make the patch better | |
| 08:42:01 | reedip | RuiChen but what about the Default Quota now | |
| 08:42:09 | reedip | I need to remove the ID from URI | |
| 08:42:11 | reedip | :) | |
| 08:43:41 | RuiChen | in default quota case, project_id is not the resource id, it's attribute | |
| 08:43:59 | RuiChen | the resource base_path is base_path = '/quotas/%(project)s/default' | |
| 08:44:17 | reedip | So how would that work ? | |
| 08:44:45 | RuiChen | uri = self.base_path % self._uri.attributes in resource class | |
| 08:44:54 | RuiChen | in _prepare_request() | |
| 08:46:31 | reedip | so you are stating that we do not need to change anything for DefaultQuotas?> | |
| 08:47:48 | RuiChen | only to change the def get_quota-default(), pass id and requires_id=False | |
| 08:48:23 | reedip | yes, thats what I was thinking | |
| 08:48:36 | reedip | so I do not need to change much for the defaultquotas patch | |
| 08:48:42 | reedip | from the defaultquotas class | |
| 08:50:02 | openstackgerrit | Reedip proposed openstack/python-openstacksdk: Fix Setting Quotas in Neutron https://review.openstack.org/419911 | |
| 08:50:19 | reedip | RuiChen vv | |
| 08:53:00 | reedip | RuiChen vv | |
| 08:53:16 | openstackgerrit | Reedip proposed openstack/python-openstacksdk: Add project ID in QuotaDefault requests https://review.openstack.org/420301 | |
| 08:54:37 | openstackgerrit | Reedip proposed openstack/python-openstacksdk: Add project ID in QuotaDefault requests https://review.openstack.org/420301 | |
| 08:57:12 | RuiChen | reedip: I had added my +1 for the two patches, thank you | |
| 08:58:14 | reedip | Thanks RuiChen :) | |
| 11:37:56 | openstackgerrit | Chris Dent proposed openstack/api-wg: [WIP] Refactor and re-validate api change guidelines https://review.openstack.org/421846 | |
| 15:48:16 | mordred | dtroyer, stevemar: did we release occ already? | |
| 15:49:09 | mordred | ah - looks like we have | |
| 15:49:16 | dtroyer | https://review.openstack.org/#/c/421469/ yes | |
| 15:49:25 | dtroyer | need another? | |
| 15:49:41 | mordred | well - we got this submitted: https://review.openstack.org/#/c/421668/ which seems both easy and friendly - but also not _super_ important | |
| 15:51:28 | dtroyer | +A, the deadline is tomorrow if we want that in stable | |
| 15:51:30 | dtroyer | wait | |
| 15:51:36 | dtroyer | o-c-c- doesn't do stable | |
| 15:52:09 | dtroyer | so its just a matter of when it would be released, today/tomorrow or next month | |
| 15:54:42 | mordred | oh - neat. cool - no rush then | |
| 15:59:26 | openstackgerrit | Dean Troyer proposed openstack/golang-client: Revert "Add crud operation of volume API v3" https://review.openstack.org/422055 | |
| 16:01:12 | ankur-gupta-f2 | dtroyer: similar note. Does OSC do a feature freeze? | |
| 16:02:59 | dtroyer | ankur-gupta-f: yes, next week. we will have one more release before then, early next week | |
| 16:03:24 | dtroyer | osc-lib freezes tomorrow, its to-become-stable-brans is the current release | |
| 16:03:33 | dtroyer | s/brans/branch/ | |
| 16:03:52 | stevemar | dtroyer: we'll have to get more strict about feature freeze and such soon | |
| 16:04:26 | dtroyer | stevemar: soon == now for the new bits (volume v3, etc). | |
| 16:08:45 | rosmaita | cdent: would be interested in your opinion of the discussion happening on https://review.openstack.org/#/c/420038/ , especially in light of your email this morning about the new tag proposal | |
| 16:09:02 | rosmaita | elmiko: ^^ etoews: ^^ | |
| 16:09:25 | ankur-gupta-f2 | dtroyer: Okay. I have told our team that any reviews coming in on any of these https://etherpad.openstack.org/p/osc-transition-priorities to be addressed ASAP. Thanks | |
| 16:09:30 | cdent | rosmaita: thanks, will look. I'm experiencing "stability" fatigue | |
| 16:09:52 | rosmaita | cdent: ha! well, i will warn you in advance, that this will not help :) | |
| 16:11:38 | elmiko | rosmaita: i'll try to take a look, i'm putting out fires at the moment :/ | |
| 16:11:50 | dtroyer | ankur-gupta-f: at this point new commands (not already in-flight) are probably doubtful | |
| 16:13:00 | rosmaita | elmiko: best wishes for your firefighting! | |
| 16:13:00 | ankur-gupta-f2 | dtroyer: most have been in flight and are nearing completion. Just waiting for the final few reviews | |
| 16:13:41 | cdent | dtroyer: yeah, I have labeling fatigue too | |
| 16:13:56 | elmiko | rosmaita: cheers! | |
| 16:15:39 | cdent | dtroyer: I tend to think of consistency on the spatial (amongst apis) dimension and some term (maybe stability) is need for the time (within an api) dimension | |
| 16:17:48 | dtroyer | cdent: they both make me tired and make me wonder why in the world this is all so hard | |
| 16:21:41 | cdent | my eyes are fuzzy enough that I ready dtroyer as saying "makes me wonder why I'm in this world" | |
| 16:22:42 | openstackgerrit | Dean Troyer proposed openstack/golang-client: Fix unit tests https://review.openstack.org/422080 | |
| 16:23:20 | cdent | I've withheld my opinion on this topic because at least for the time being it is probably too poorly informed to be useful, but yeah it all makes me pretty tired | |
| 16:26:26 | dtroyer | here is mine: one place that a BDFL is useful is in keeping at least some of the cats in the same order everywhere. I've had to say no so many times to OSC command stuff just to be as close to consistent as we are (still a ways to go). There is no way in h3ll that we'd have accomplished that with teams doing their own thing, even if they agreed on the idea. Even now, many of the OSC plugins slide past the guidelines (I d | |
| 16:26:36 | dtroyer | That is just the CLI, and "CLI's are easy" | |
| 16:27:46 | dtroyer | the only way we will get a clean, consistent REST API for OpenStack is to have a single unit (team, working group, whatever) build it <cough>oaktree</cough> | |
| 16:28:08 | dtroyer | </rant> | |
| 16:28:49 | cdent | I can get behind that (modulo corporate enterprisey nonsense) | |
| 16:30:05 | dtroyer | I think it would take a corporate-type sponsor without an API to defend/enhance to sponsor it to ever get anywhere | |