05:00:53 #startmeeting servicevm-device-manager 05:00:54 Meeting started Tue Aug 19 05:00:53 2014 UTC and is due to finish in 60 minutes. The chair is yamahata. Information about MeetBot at http://wiki.debian.org/MeetBot. 05:00:55 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 05:00:57 The meeting name has been set to 'servicevm_device_manager' 05:01:23 #topic Announcement 05:01:34 I have no special announcement today 05:01:41 Does anyone have any? 05:02:47 Hi all. Nothing from me. At the moment busy getting stuff done for J3. 05:03:00 bmelande: hi 05:03:01 Neutron feature freeze is this Thursday 05:03:20 Sure. it's busy now. 05:03:31 #open discussion 05:03:47 For now I'm addressing config agent and plugging drivers. 05:03:59 This week, I'll upload WIP code to github 05:04:18 nice 05:04:31 I think patch review is urgent issue? 05:04:51 Which patch? 05:05:01 bmelande: yours and Karthik 05:05:12 Aha. :-) 05:05:27 #link https://review.openstack.org/#/c/101002/ 05:05:40 #link https://review.openstack.org/#/c/102336/ 05:06:11 I posted a simplification patch for csr1kv. Does it make sense? 05:06:13 hi 05:06:38 #link https://review.openstack.org/#/c/115157/ csr1kv simplification 05:06:40 natarajk: hi 05:07:24 bmelande: or do you have any plan to use locally created file by hosting device driver? 05:07:34 yamahata: Yes, it was good. Thanks. The configdrive change you pointed out is also useful. Has adopted that. Working on the 3rd party CI. That's why no new upload yet. 05:08:15 bmelande: Great, then I'll abandon mine. 05:08:24 yamahata: No, I'm happy to not have to generate a file and later remove it. Just a hazzle. 05:08:54 bmelande: okay 05:09:25 bmelande: I have another question. Is complementary_id necessary? Doesn't id work? 05:12:20 bmelande: Maybe a small patch to show the idea would help. 05:13:07 yamahata: I had to add that to ensure cleaning up the ports/networks I create is alwoays possible to remove. 05:13:53 bmelande: Oh, I'll check it from that point of view. 05:14:14 yamahata: you kind of already have it 05:14:49 maybe, I'll try to check it by creating a patch 05:15:29 yahamata: because you had this attribute to hold the uuid of the Nova VM 05:16:47 I'm fine with keeping complementary_id in table 05:17:16 yamahata: Actually, I like the name you have used better. :-) 05:19:18 anything else to discuss? 05:19:24 bmelande: The problem I had was that I set the id of the hosting device to the Nova VM's uuid. But if spawning fails when Nova has been given the instruction to spin up VM, *but* before the ports for it have been marked with that uuid, I had no good way of identifying theose ports for a later clean up. 05:20:40 bmelande: I see. It's error recovery path. I totally forgot it as I've wanted to have something working first. 05:20:55 yamahata: Sorry, last was for you. That was the reasoning why I added it. But as said, afaik, you store the Nova uuid in a separate attribute and not in the hosting device id so you pretty miuch have the complementeary id already. 05:21:50 I understood its background. 05:22:28 thanks for the explanation 05:23:53 natarajk: s3wong do you have any topics? 05:24:11 yamahata: no, all good for me 05:24:18 have been busy with our CI setup to run all the required tests 05:24:37 hopefully i can contribute more from next week 05:24:45 natarajk: You too. :-) 05:24:52 natarajk: No problem. good luck for your CI 05:25:00 so seems no more topics. 05:25:15 thank you every one. see you next week. 05:25:30 thanks 05:25:36 Thanks. Bye 05:25:40 bye 05:25:40 #endmeeting