| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-05-31 | |||
| 08:19:58 | openstackgerrit | jichenjc proposed openstack/nova master: [placement] Add allocation doc for placement https://review.openstack.org/450550 | |
| 08:28:20 | openstackgerrit | jichenjc proposed openstack/nova master: make route and controller in alpha sequence https://review.openstack.org/469380 | |
| 08:31:49 | Kevin_Zheng | gmann: thanks | |
| 08:38:25 | openstackgerrit | Huan Xie proposed openstack/nova master: XenAPI: Remove bittorrent.py which is already deprecated https://review.openstack.org/468273 | |
| 08:48:02 | markus_z | trivial fix: https://review.openstack.org/#/c/469365/2 (which unblocks the docs gate) | |
| 08:50:22 | openstackgerrit | jichenjc proposed openstack/nova master: Use plain routes list for hosts endpoint instead of stevedore https://review.openstack.org/459231 | |
| 08:50:22 | openstackgerrit | jichenjc proposed openstack/nova master: use plain routes list for hypervisor endpoint instead of stevedore https://review.openstack.org/459235 | |
| 09:26:25 | bhagyashris | johnthetubaguy_ : Hi, Could you please take some insights for patch: https://review.openstack.org/#/c/409644/ | |
| 09:27:01 | openstackgerrit | falseuser proposed openstack/nova master: Clear instance root_device_name in rebuild operation https://review.openstack.org/464991 | |
| 09:28:48 | egonzalez | Hi, seing a bunch of this error on nova-api logs regarding metadata PicklingError: Can't pickle <class 'sqlalchemy.orm.session.Session'>: it's not the same object as sqlalchemy.orm.session.Session | |
| 09:29:06 | egonzalez | is there a known issue in master, full trace http://paste.openstack.org/show/611064/ | |
| 09:34:05 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add PCIWeigher https://review.openstack.org/379524 | |
| 09:34:06 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Prefer non-PCI host nodes for non-PCI instances https://review.openstack.org/379625 | |
| 09:35:56 | sfinucan | egonzalez: I can't say I've seen it, no. Have you opened a bug? | |
| 09:36:31 | egonzalez | sfinucan, nope, just wanted to make sure if is a known issue before | |
| 09:36:51 | sfinucan | egonzalez: Ah, in that case...not that I'm aware of, no :) | |
| 09:37:05 | egonzalez | sfinucan, going to retrieve more info and open a bug | |
| 09:48:20 | alex_xu | Kevin_Zheng: gmann done | |
| 09:49:10 | Kevin_Zheng | cool thx | |
| 10:00:11 | openstackgerrit | Merged openstack/nova master: Fix doc job with correct ref link https://review.openstack.org/469365 | |
| 10:00:58 | andymccr | egonzalez: we're seeing similar errors for memcache - started happening on the 23rd of may. | |
| 10:03:22 | egonzalez | andymccr, thanks will open a bug with more info, same errors are found in kolla gates | |
| 10:03:41 | andymccr | egonzalez: can you link the bug once it's created - i'll add our logs too | |
| 10:03:50 | openstackgerrit | Huan Xie proposed openstack/nova master: XenAPI: Remove bittorrent.py which is already deprecated https://review.openstack.org/468273 | |
| 10:13:32 | egonzalez | andymccr, https://bugs.launchpad.net/nova/+bug/1694666 | |
| 10:13:32 | openstack | Launchpad bug 1694666 in OpenStack Compute (nova) "metadata service PicklingError: Can't pickle <class 'sqlalchemy.orm.session.Session'>" [Undecided,New] | |
| 10:17:26 | andymccr | egonzalez: thanks! | |
| 10:40:41 | sfinucan | alex_xu: You might look at https://review.openstack.org/#/c/399613/ today. It's the only patch in that series that still needs a +2, and you've already reviewed the rest | |
| 10:58:53 | openstackgerrit | Sean Dague proposed openstack/nova master: Send request_id on neutron calls https://review.openstack.org/468356 | |
| 11:11:09 | openstackgerrit | Merged openstack/nova master: Added nova objects for intance diagnostics https://review.openstack.org/434745 | |
| 11:11:41 | openstackgerrit | Merged openstack/nova master: Replace diagnostics objects with Nova diagnostics objects https://review.openstack.org/394480 | |
| 11:43:23 | bauzas | cdent: edleafe: jaypipes: sfinucan: stupidly forgot that I was normally on PTO today, so I won't be there today | |
| 11:50:29 | openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Send out notifications when instance tags changed https://review.openstack.org/459923 | |
| 12:13:57 | openstackgerrit | Merged openstack/nova master: XenAPI: nova-compute cannot restart after manually delete VM https://review.openstack.org/467926 | |
| 12:14:45 | openstackgerrit | Merged openstack/nova master: Changing deleting stale allocations warning to debug https://review.openstack.org/468519 | |
| 12:20:20 | dims | bauzas : easy one if you get a chance please peek - https://review.openstack.org/#/c/468743/ | |
| 12:27:42 | sfinucan | dims: Is https://bugs.launchpad.net/nova/+bug/1694666 in any way related to that? | |
| 12:27:42 | openstack | Launchpad bug 1694666 in OpenStack Compute (nova) "metadata service PicklingError: Can't pickle <class 'sqlalchemy.orm.session.Session'>" [Undecided,New] | |
| 12:27:48 | sfinucan | that being | |
| 12:27:51 | sfinucan | https://review.openstack.org/#/c/468743/ | |
| 12:29:10 | sfinucan | I +W'd that, given that sdague's comments were address and it otherwise looked a-ok | |
| 12:31:38 | dims | sfinucan : thanks! no it was a badly written test (not the Pickling error thingy) | |
| 12:35:30 | openstackgerrit | Hong Hui Xiao proposed openstack/nova master: VMware: Check if network is ready on cluster host https://review.openstack.org/467589 | |
| 12:37:31 | openstackgerrit | Gábor Antal proposed openstack/nova master: Make docstring comments more consistent in rpcapi.py https://review.openstack.org/380344 | |
| 12:38:59 | alex_xu | sfinucan: yea, I will try | |
| 12:39:14 | sfinucan | alex_xu: Ta | |
| 12:39:55 | alex_xu | sfinucan: also an easy one to get rid of the useless warning in the test https://review.openstack.org/#/c/468614/ | |
| 12:41:27 | alex_xu | Kevin_Zheng: I still didn't see the last patch for this series https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/support-tag-instance-when-boot, is the last patch in the same topic? | |
| 12:43:32 | Kevin_Zheng | alex_xu: thanks for asking, I was quite busy recently and didn't got chance to update it | |
| 12:44:17 | Kevin_Zheng | But since the 3rd one is almost Ok, I will update the 4th one tomorrow | |
| 12:45:27 | alex_xu | Kevin_Zheng: cool, thanks, no worries, just want to ensure i'm not missing something | |
| 12:45:41 | alex_xu | nova api meeting is in 15 mins | |
| 12:47:25 | sfinucan | alex_xu: Done and dusted | |
| 12:48:34 | alex_xu | sfinucan: thanks | |
| 12:52:52 | edleafe | jaypipes: sorry, didn't have much time yesterday between meetings and such | |
| 12:53:04 | edleafe | jaypipes: what is the status of that odd os-traits bug? | |
| 13:00:27 | alex_xu | nova api meeting is running at #openstack-meeting-4 | |
| 13:24:08 | cdent | edleafe: I'm reviewing in the https://review.openstack.org/#/q/topic:bp/placement-claims-4 stack. Is it just the first two that matter for now? | |
| 13:24:46 | cdent | and what of the older topic matters: https://review.openstack.org/#/q/topic:bp/placement-claims | |
| 13:26:24 | openstackgerrit | Zhenyu Zheng proposed openstack/nova master: Send out notifications when instance tags changed https://review.openstack.org/459923 | |
| 13:30:07 | jaypipes | edleafe: have not looked at it since yesterday | |
| 13:30:40 | jaypipes | efried: awesome review on the provider tree patch. was working into the wee hours fixing that up. just rebasing to master now and will push a new patch shortly. | |
| 13:39:57 | edleafe | cdent: those two are the current focus. After getting feedback that operators are indeed (mis-)using the host_subset_size, I need to remove that change from the stack | |
| 13:40:31 | cdent | edleafe: k, I'll not bother with the rest for now then, moving on to os-traits | |
| 13:40:35 | cdent | thanks | |
| 13:41:46 | openstackgerrit | Merged openstack/nova master: Do not rely on dogpile internals for mocks https://review.openstack.org/468743 | |
| 13:44:32 | cdent | jaypipes: left some comments on the traits syncing stuff. I don't understand why need a cache at all? | |
| 13:46:11 | jaypipes | cdent: it was a convenient place to put the sync routine and it matches the resource class cache. | |
| 13:46:43 | cdent | sure, but the resource class cache is needed because we have some stuff in db, and some stuff not in db | |
| 13:46:46 | cdent | not true with traits | |
| 13:46:56 | jaypipes | yes, I'm aware of that | |
| 13:47:07 | cdent | so...? | |
| 13:47:09 | jaypipes | cdent: it was a convenient place to put the sync routine and it matches the resource class cache. | |
| 13:47:15 | jaypipes | cdent: it was a convenient place to put the sync routine and it matches the resource class cache. | |
| 13:47:16 | jaypipes | cdent: it was a convenient place to put the sync routine and it matches the resource class cache. | |
| 13:47:28 | cdent | yeah, that's completely meaningless :) | |
| 13:47:32 | jaypipes | lol, sorry :) | |
| 13:47:35 | cdent | you can say it ten more times | |
| 13:47:41 | jaypipes | didn't mean to repeat that. | |
| 13:47:59 | cdent | :) | |
| 13:48:03 | jaypipes | serves me right for trying to do two things at once :) | |
| 13:48:16 | cdent | only two? that's pretty good discipline | |
| 13:49:01 | jaypipes | cdent: anyway, did you see the original patch from one of alex_xu's colleagues that added a new placement-manage CLI tool to do trait syncs? | |
| 13:49:11 | cdent | yes | |
| 13:49:23 | cdent | we don't want that | |
| 13:49:24 | jaypipes | cdent: so you're aware of the impetus for me doing this, then. | |
| 13:49:42 | jaypipes | cdent: how do you propose to do the syncing? | |
| 13:49:57 | cdent | I'm aware of why you want to have a sync happen in an automatic fashion, but it ought to be possible to do it without the cache | |
| 13:50:03 | cdent | one sec | |
| 13:50:26 | mriedem | sfinucan: i added cfriesen to your series here https://review.openstack.org/#/q/topic:bp/reserve-numa-with-pci - i'd like his review on that before i do | |
| 13:50:41 | mriedem | moshe would probably also be good | |
| 13:50:58 | jaypipes | cdent: oh, it's totally possible to do it without the cache. I just figured the trait cache would be useful in the same way the rc cache is | |
| 13:51:11 | cdent | jaypipes: that's my question: what's the point of the cache? | |
| 13:51:37 | cdent | " I don't understand why need a cache at all?" | |
| 13:51:38 | jaypipes | cdent: just consistency with the resource classes and a convenient place to put the sync thing, like I said. | |
| 13:51:47 | jaypipes | cdent: consistency with rcs | |
| 13:52:04 | jaypipes | cdent: where would you put the auto-sync thing? | |
| 13:52:30 | jaypipes | cdent: feel free to push a patch that doesn't use a cache. I'm happy to review it. | |
| 13:52:52 | cdent | instead of _ensure_trait_cache have _ensure_trait_sync which checks a global boolean to know if sync happend, get rid of all the code in the trait cache file except for the sync method | |
| 13:54:13 | cdent | jaypipes: I'll be happy to push something, but just as a point of order, I don't think it should be incumbent on me to prove why you shoudn't add a piece of code that doesn't appear necessary. It should be on you to prove that it is (beyond consistency). | |