=== giraffe is now known as Guest84199 === pavlushka is now known as Guest51462 === Guest51462 is now known as pavlushka [06:39] Hello [06:41] can we assign more than one ip on AWS EC2 dedicated server [06:44] jamespage: seems neutron rc3 is needed https://bugs.launchpad.net/bugs/1628549 [06:44] Launchpad bug 1628549 in neutron "DB migration is broken with two unassigned floating IPs" [High,Fix released] [06:45] can we assign more than one ip on AWS EC2 dedicated server ? [07:56] Good morning. [08:05] Hi, I have a dedicated ubuntu server and I want to run a clean install on it and migrate all data back to it. Would it be easy to request an image and just install that image on a local PC here as backup or would I run into hardware problems then? [08:06] I think easiest would be buying a new dedicated server and cancelling this one after migration, but their prices seem to have gone up a lot through the years [08:25] is it common for NFS to cause high load averages with minimal cpu? [08:36] thekrynn: Depends on the usage, I'd say. If a hundreds of clients are hammering the NFS server, yeah expect a high load. [08:36] lordievader: basically i have an NFS server, an NFS client, both running on the same hypervisor, and the client is simply grabbing data, awk'ing it, and then writing back small files [08:36] at around 500Mbit/s [08:38] Check vmstat I'd say, if you see a lot of blocked processes and a high io-wait time, might be the cause. [09:12] thanks lordievader [09:12] seems like cating the data over nfs into a pipe was causing a lot of the issues [09:12] i changed that out and the cpu wait in vmstat went down considerably [09:15] ;) [09:15] Glad you solved it. === MrBIOS_ is now known as MrBIOS === Malediction_ is now known as Malediction === _degorenko is now known as degorenko [11:37] jamespage: coreycb: could you check for working dependencies here, please? https://bugs.launchpad.net/keystone/+bug/1628883 [11:37] Launchpad bug 1628883 in OpenStack Identity (keystone) "Minimum requirements too low on oslo.log for keystone" [Undecided,New] [11:41] frickler, we'll fix that up, thanks for reporting it. [11:43] coreycb: thanks, I'll also go and see what upstream thinks about this [11:46] frickler, it seems that global-requirements is too low: https://github.com/openstack/requirements/blob/stable/newton/global-requirements.txt [11:47] frickler, keystone has a higher min version in their requirements.txt but it should align with g-r [11:47] frickler, I'll add upstream to the bug [11:50] frickler, nm they are aligned at >= 1.14.0. I was on the wrong keystone branch. [11:51] still a problem obviously === Mobutils_ is now known as Mobutils [12:53] frickler, jamespage, ddellav: I just took a pass on all of our core newton packages to bump oslo.log >= 3.16.0 [14:42] rbasak: nacc: jgrimm: regarding the current clamav package in Yakkety, I don't think that we should wait on the MIR for tomsfastmath [14:43] mdeslaur has uploaded 0.99-2 in all the stable releases while keeping the old in-package library so I think that Yakkety should to the same [14:43] then we can wait for 17.04 to MIR the library [14:49] caribou, that works for me [14:51] caribou, though.. its a bit odd that the MIR should take long?? that is, isn't it just a breakout of the previously embedded library which is already in main. [14:52] caribou, but, i'm find with that given where we are at in the cycle [14:52] jgrimm: the security team already has a long backlog of MIR reviews [14:52] fine === lutostag_ is now known as lutostag === JanC is now known as Guest12605 === JanC_ is now known as JanC [14:53] caribou, I added it here for tracking -> https://blueprints.launchpad.net/ubuntu/+spec/servercloud-z-server-core [14:53] thanks [15:32] ddellav, gnocchi ok to sync? [15:32] ddellav, and magnum? [15:33] coreycb those were a bit weird because they aren't RC1, they are full version releases. I wasn't sure if they were applicable. I will run quick builds on them now. [15:34] jamespage: can't recall if i saw, but did you have a test import you wanted us to run for the git workflow? [15:34] ddellav, they've likely just made it to final release already [15:42] coreycb gnocchi needs a python-gabbi sync. gabbi builds in xenial and yakkety without issue. [15:44] ddellav, ok sync initiated for gabbi [15:55] coreycb how are we doing with python-os-api-ref? That's needed for magnum as well [15:55] as is python-k8sclient which builds on both fine and could use a sync [15:56] coreycb, quick poke [15:56] nova -> os-brick requires privsep [15:57] ddellav, looks like os-api-ref is in the archive now [15:57] ddellav, at least, it's not in the NEW queue anymore https://launchpad.net/ubuntu/yakkety/+queue?queue_state=1&queue_text= [15:58] ddellav, eh... wrong queue, still there: https://launchpad.net/ubuntu/yakkety/+queue?queue_state=0&queue_text= [15:59] ddellav, I poked the release team this morning about it, I'll poke again [15:59] jamespage, ok, yeah it looks like we can't get around it [16:02] jamespage, I'm following up on openstack VMT to ensure it's security supported by upstream [16:02] coreycb, +1 [16:02] sarnold, ^^ [16:07] ddellav, in the mean time can you just build anything that depends on os-api-ref in a ppa and we'll get the syncs in the queue? [16:10] coreycb yep [16:10] ddellav, thanks [16:11] rockstar, pylxd 2.1.1 uploaded to yakkety [16:34] coreycb: aces [16:48] ddellav, k8sclient sync initiated === Jare_ is now known as Jare === pavlushka is now known as Guest94932 === degorenko is now known as _degorenko === Guest94932 is now known as pavlushka [18:29] jamespage,coreycb, thanks (openstack vmt supporting oslo.privsep) [18:30] sarnold, you're welcome, thanks for the review === blizzow1 is now known as blizzow === ddstreet_away is now known as ddstreet === andol_ is now known as andol [20:19] jamespage, ddellav: I opened bug 1629097 for neutron in newton [20:19] bug 1629097 in neutron "neutron-rootwrap processes not getting cleaned up" [Undecided,New] https://launchpad.net/bugs/1629097 [20:20] seems to be the cause of memory exhaustion in our charm deploys of nova-compute and neutron-gateway [20:20] beisner, ^ [20:30] coreycb, broken dns? [20:31] jamespage, think so? [20:31] jamespage, I deployed mitaka for comparison and there weren't any issues [20:32] jamespage, and reverting that commit seemed to help. let me unrevert to double check. [20:33] maybe not [20:45] jamespage, ok confirmed that reverting that commit fixes it [20:57] coreycb, awesome === nacc_ is now known as nacc [23:16] dannf, hey - I think that arm64 hugepages patch is a little late for newton fwiw [23:16] esp as its not accepted upstream yet