*** rebase has quit IRC | 00:21 | |
*** rebase has joined #openstack-operators | 00:22 | |
*** sticker has joined #openstack-operators | 00:32 | |
*** vinsh has quit IRC | 00:34 | |
*** vinsh has joined #openstack-operators | 00:35 | |
*** arif-ali has quit IRC | 00:45 | |
*** dalees has quit IRC | 00:45 | |
*** dalees has joined #openstack-operators | 00:46 | |
*** arif-ali has joined #openstack-operators | 00:46 | |
*** Apoorva_ has joined #openstack-operators | 00:49 | |
*** Apoorva_ has quit IRC | 00:49 | |
*** Apoorva_ has joined #openstack-operators | 00:50 | |
*** Apoorva_ has quit IRC | 00:52 | |
*** Apoorva has quit IRC | 00:52 | |
*** zerick has quit IRC | 00:55 | |
*** zerick has joined #openstack-operators | 00:57 | |
*** haukebruno has quit IRC | 01:00 | |
*** keekz has quit IRC | 01:00 | |
*** keekz has joined #openstack-operators | 01:00 | |
*** haukebruno has joined #openstack-operators | 01:01 | |
*** tdasilva has quit IRC | 01:21 | |
*** tdasilva has joined #openstack-operators | 01:25 | |
*** mriedem has quit IRC | 01:28 | |
*** haukebruno has quit IRC | 01:29 | |
*** haukebruno has joined #openstack-operators | 01:34 | |
*** fragatin_ has quit IRC | 01:57 | |
*** fragatina has joined #openstack-operators | 01:58 | |
*** dtrainor has quit IRC | 02:16 | |
*** rebase has quit IRC | 02:21 | |
*** rebase has joined #openstack-operators | 02:22 | |
*** B_Smith has quit IRC | 02:23 | |
*** rebase has quit IRC | 02:28 | |
*** cemason has joined #openstack-operators | 03:06 | |
*** rebase has joined #openstack-operators | 03:28 | |
*** rebase has quit IRC | 03:33 | |
*** liverpooler has quit IRC | 03:43 | |
*** dtrainor has joined #openstack-operators | 03:51 | |
*** jamemcc has quit IRC | 04:09 | |
*** JillS has quit IRC | 04:10 | |
*** arnewiebalck_ has joined #openstack-operators | 04:13 | |
*** fragatin_ has joined #openstack-operators | 04:22 | |
*** zenirc369 has joined #openstack-operators | 04:22 | |
*** fragatina has quit IRC | 04:26 | |
*** fragatin_ has quit IRC | 04:27 | |
*** fragatina has joined #openstack-operators | 04:45 | |
*** fragatina has quit IRC | 04:45 | |
*** fragatina has joined #openstack-operators | 04:46 | |
*** JillS has joined #openstack-operators | 04:56 | |
*** yprokule has joined #openstack-operators | 04:57 | |
*** jamemcc has joined #openstack-operators | 04:57 | |
*** slaweq has joined #openstack-operators | 05:01 | |
*** slaweq has quit IRC | 05:02 | |
*** arnewiebalck_ has quit IRC | 05:20 | |
*** sticker_ has joined #openstack-operators | 05:25 | |
*** slaweq has joined #openstack-operators | 05:26 | |
*** darrenc has quit IRC | 05:28 | |
*** darrenc has joined #openstack-operators | 05:28 | |
*** slaweq has quit IRC | 05:28 | |
*** sticker has quit IRC | 05:29 | |
*** rebase has joined #openstack-operators | 06:10 | |
*** Oku_OS-away is now known as Oku_OS | 06:11 | |
*** zenirc369 has quit IRC | 06:16 | |
*** racedo has joined #openstack-operators | 06:26 | |
*** d0ugal has quit IRC | 06:39 | |
*** d0ugal has joined #openstack-operators | 06:41 | |
*** makowals has quit IRC | 06:44 | |
*** makowals has joined #openstack-operators | 06:45 | |
*** manheim has joined #openstack-operators | 07:00 | |
*** racedo has quit IRC | 07:02 | |
*** tesseract has joined #openstack-operators | 07:04 | |
*** manheim has quit IRC | 07:04 | |
*** manheim has joined #openstack-operators | 07:04 | |
*** pcaruana has joined #openstack-operators | 07:19 | |
*** robksawyer has joined #openstack-operators | 07:29 | |
*** robksawyer has quit IRC | 07:33 | |
*** electrofelix has joined #openstack-operators | 07:34 | |
*** robksawyer has joined #openstack-operators | 07:42 | |
*** arnewiebalck_ has joined #openstack-operators | 07:45 | |
*** robksawyer has quit IRC | 07:46 | |
*** dbecker has joined #openstack-operators | 08:00 | |
*** rebase has quit IRC | 08:04 | |
*** rebase has joined #openstack-operators | 08:06 | |
*** arnewiebalck_ has quit IRC | 08:10 | |
*** yprokule has quit IRC | 08:17 | |
*** racedo has joined #openstack-operators | 08:30 | |
*** sticker_ is now known as sticker | 08:33 | |
*** derekh has joined #openstack-operators | 08:36 | |
*** rebase has quit IRC | 08:53 | |
*** aojea has joined #openstack-operators | 08:55 | |
*** aojea_ has joined #openstack-operators | 09:19 | |
*** aojea has quit IRC | 09:23 | |
*** haplo37 has quit IRC | 09:47 | |
*** haplo37 has joined #openstack-operators | 09:56 | |
*** paramite has joined #openstack-operators | 10:05 | |
*** B_Smith has joined #openstack-operators | 10:09 | |
*** slaweq has joined #openstack-operators | 10:29 | |
*** slaweq has quit IRC | 10:34 | |
*** JillS has quit IRC | 10:37 | |
*** jamemcc has quit IRC | 10:38 | |
*** yprokule has joined #openstack-operators | 10:57 | |
*** Miouge has joined #openstack-operators | 11:06 | |
*** JillS has joined #openstack-operators | 11:16 | |
*** jamemcc has joined #openstack-operators | 11:16 | |
*** vijaykc4 has joined #openstack-operators | 12:30 | |
*** vijaykc4 has quit IRC | 12:32 | |
*** vijaykc4 has joined #openstack-operators | 12:33 | |
*** liverpooler has joined #openstack-operators | 12:54 | |
*** mriedem has joined #openstack-operators | 13:03 | |
*** VW has joined #openstack-operators | 13:04 | |
*** jamesdenton has joined #openstack-operators | 13:16 | |
*** tongli has joined #openstack-operators | 13:20 | |
*** vijaykc4 has quit IRC | 13:21 | |
*** Miouge has quit IRC | 13:21 | |
*** fragatina has quit IRC | 13:40 | |
*** fragatina has joined #openstack-operators | 13:41 | |
*** catintheroof has joined #openstack-operators | 13:56 | |
*** chlong has joined #openstack-operators | 13:58 | |
*** cnf has quit IRC | 14:04 | |
*** Miouge has joined #openstack-operators | 14:14 | |
*** slaweq has joined #openstack-operators | 14:30 | |
*** slaweq has quit IRC | 14:34 | |
*** manheim has quit IRC | 14:35 | |
*** manheim has joined #openstack-operators | 14:40 | |
*** manheim has quit IRC | 14:46 | |
*** marst has quit IRC | 14:47 | |
*** manheim has joined #openstack-operators | 14:48 | |
wasmum | moving from kilo to mitaka some quick questions: We are using network nodes on kilo with vrrp | 14:48 |
---|---|---|
wasmum | how does this look in mitaka going to linux bridging? Do you still need your network node, I assume yes. How does this effect the controller build, does it just get the neutron server service api installed and you build your network nodes as you would build a controller node (acccording to the docs) without the neutron server service? | 14:50 |
*** manheim has quit IRC | 14:53 | |
*** marst has joined #openstack-operators | 14:54 | |
*** manheim has joined #openstack-operators | 14:55 | |
*** marst_ has joined #openstack-operators | 14:55 | |
*** marst has quit IRC | 14:55 | |
*** manheim has quit IRC | 14:57 | |
*** tongli has quit IRC | 15:00 | |
*** cemason has quit IRC | 15:07 | |
*** cemason has joined #openstack-operators | 15:09 | |
*** DaSchab has joined #openstack-operators | 15:09 | |
*** catintheroof has quit IRC | 15:10 | |
*** catintheroof has joined #openstack-operators | 15:11 | |
*** DaSchab has quit IRC | 15:11 | |
*** catintheroof has quit IRC | 15:11 | |
*** Miouge has quit IRC | 15:12 | |
*** aojea_ has quit IRC | 15:29 | |
*** fandi has joined #openstack-operators | 15:30 | |
*** slaweq has joined #openstack-operators | 15:32 | |
*** rmart04 has joined #openstack-operators | 15:37 | |
*** chyka has joined #openstack-operators | 15:39 | |
*** yprokule has quit IRC | 15:43 | |
*** Rockyg has joined #openstack-operators | 15:45 | |
erhudy | moving from nova-network to neutron? | 15:53 |
*** Caterpillar has joined #openstack-operators | 15:56 | |
*** jlk has quit IRC | 16:01 | |
*** jlk has joined #openstack-operators | 16:02 | |
*** jlk has quit IRC | 16:02 | |
*** jlk has joined #openstack-operators | 16:02 | |
*** slaweq has quit IRC | 16:04 | |
*** rebase has joined #openstack-operators | 16:18 | |
*** shasha___ has joined #openstack-operators | 16:21 | |
*** Apoorva has joined #openstack-operators | 16:22 | |
wasmum | erhudy: moving from kilo to mitaka neutron networking in kilo with vrrp an gre overlay and going to mitaka linux bridging with a vxlan overlay and vrrp | 16:26 |
*** fragatina has quit IRC | 16:38 | |
erhudy | ah, godspeed you, the godaddy people in here have mentioned repeatedly that they want to jettison their OVS stuff in favor of linuxbridge so maybe one of them would have some Top Tips if they've begun moving down that path | 16:39 |
*** rebase has quit IRC | 16:42 | |
*** rebase has joined #openstack-operators | 16:42 | |
mdorman | erhudy: wasmum yeah don’t really have any practical advice at this point. that’s still a “it would be nice to do” project on our end. The biggest challenge for us would be figuring out how to migrate from OVS to linux bridge with minimal impacts to VM connectivity | 16:45 |
wasmum | mdorman: can you disclose how many workloads you average (instance count)? | 16:46 |
*** derekh has quit IRC | 16:46 | |
wasmum | at this point were under 200 instances so the impact exists but can be mitigated through politics | 16:47 |
mdorman | like VMs per hypervisor? | 16:47 |
mdorman | we have on the order of 10s of thousands of instances total | 16:47 |
wasmum | so moving from ovs to linux bridging is something we need to either commit to now or never. | 16:47 |
mdorman | yeah now would be the time to do it for sure :) | 16:47 |
erhudy | <requisite shilling for calico goes here> | 16:47 |
wasmum | yeah the impact you face would be a resume generating event | 16:48 |
mdorman | haha | 16:48 |
*** usertm has joined #openstack-operators | 16:48 | |
mdorman | taht’s why we’d have to figure out how to do it in a non impactful way, or with a blip of just seconds | 16:48 |
erhudy | moving baseline network technologies without user impact seems like a sisyphean task | 16:49 |
mdorman | agreed. i am not sure it will ever happen to be honest | 16:49 |
wasmum | sounds like an awesome challenge. not sure the grey mass in my head would be able to comprehend how to accomplish that | 16:49 |
erhudy | i spent about a year researching whether we could jump from nova-network to anything neutron in place before concluding "not without making everyone really mad" | 16:50 |
wasmum | man how I long for an LTS supported version of openstack | 16:50 |
erhudy | you and every other operator | 16:50 |
*** tesseract has quit IRC | 16:51 | |
wasmum | yeah, how come that falls silent in the foundation meetings? | 16:51 |
*** marst_ has quit IRC | 16:52 | |
erhudy | ¯\_(ツ)_/¯ | 16:52 |
wasmum | upgrades and architectural changes are the main obstacles we face presenting a long term solution in an enterprise | 16:52 |
erhudy | it's why we were on kilo for a long time after EOL, moved to liberty after it was EOL, and will be moving our existing clusters to mitaka even though it's EOL | 16:53 |
erhudy | most operators in here probably have similar stories | 16:53 |
erhudy | if you're willing to pay for canonical support, mitaka was part of the xenial LTS release so in a certain sense mitaka is LTS, but supported by canonical | 16:55 |
Rockyg | I think the problem with LTS is that the vendors aren't willing to agree on it and make it a priority for their community developers. | 17:02 |
Rockyg | Learning how to steer community developers is really hard, I think. | 17:04 |
erhudy | the lack of an official upstream LTS combined with a 6-month dev cycle means the changes arrive at a bewildering pace | 17:04 |
*** fragatina has joined #openstack-operators | 17:05 | |
Rockyg | Very few devs have ever experienced trying to run/mange a cloud | 17:05 |
Rockyg | They don't feel that pain. | 17:05 |
erhudy | well, there are a number of devs in here now and i don't want to craptalk them | 17:06 |
Rockyg | A number of senior community devs still don't understand why ops don't do continuous deployment | 17:06 |
erhudy | that particular thought has come to mind, but having been on both sides, as a dev you think about different concerns | 17:07 |
Rockyg | It's not that they're bad. It's just how do you get them to understand where the code issues intersect the ops pain points. | 17:07 |
Rockyg | They hear the pain, but they need to be aqble to drill down to specific lines of code. | 17:08 |
Rockyg | And ops don't drill to the code level. | 17:08 |
erhudy | devstack is now outlawed, every dev has to build their code on a 200-node cluster that will disintegrate under the volume of RPC messages and DB connections | 17:08 |
Rockyg | It's the different fram of references and finding the intersectiing places isn't easy. | 17:09 |
Rockyg | Uh, erhudy is that so they solve your problem first ;-) | 17:09 |
erhudy | idk, i think that handwaves over the capabilities of ops people, the old stereotype of being a "sysop" who has a pile of tribal knowledge shell scripts doesn't really apply these days | 17:09 |
Rockyg | No, the ops guys know, the dev guys know, but they work on totally different types of systems | 17:10 |
erhudy | as somebody with an ops background, i do think that devs need to walk that walk at least once | 17:11 |
erhudy | in order to understand why your ops guys are telling you what they are, instead of assuming they're stonewalling because lazy/antagonistic | 17:12 |
Rockyg | And maybe devops has so much code themselves that devs aren't aware of, it would take some time for a dev to review the ops code base to understand the interoperations between the two code bases. | 17:12 |
*** paramite has quit IRC | 17:12 | |
Rockyg | Nah. The devs don't think ops are lazy. It's just two different languages with the same words | 17:12 |
erhudy | we'll have the chance to have our own west side story week after next | 17:13 |
erhudy | devs are jets, operators are sharks | 17:13 |
erhudy | brawl in the convention center, total madness | 17:14 |
Rockyg | lol | 17:14 |
Rockyg | I've been promoting a take a dev to work day to the ops folks I know, but it hasn't gained much steam. | 17:15 |
Rockyg | The OSIC guys actually have a setup where the devs get a week of ops tasks to learn and do before they start contributing to openstack code | 17:16 |
*** cnf has joined #openstack-operators | 17:18 | |
Rockyg | also, with the novanet to neutron move, there were plans and work on smooth migration, but it took so long to actually get the code there that lots of that energy and direction got left on the roadside | 17:18 |
erhudy | re: that, our specific problem was that we use nova-network in multi-host mode, so our clusters never had network nodes | 17:19 |
erhudy | and DVR wasn't in the right place at the time, and you still need network nodes to do routing for instances without provider network IPs anyway | 17:19 |
Rockyg | Oh. that. | 17:19 |
erhudy | calico is what we ended up settling on since it's a private cloud | 17:20 |
*** aojea has joined #openstack-operators | 17:20 | |
Rockyg | Uh, yeah. Devs can't code to architectures they don't know about/expect. | 17:20 |
erhudy | it's not like we hacked multi-host mode into nova-network, it was a flag in nova.conf | 17:21 |
Rockyg | I bet you could get Neil Gerram of calico to work through some upgrade scenarios with you at the summit. | 17:21 |
Rockyg | Jerram | 17:21 |
erhudy | we already decided our existing nova-network clusters will get upgraded to mitaka and then they can live out the rest of their lives in peace | 17:22 |
Rockyg | Grab him if he's there and and do some brainstorming. He'd be good at that and would even enjoy it. | 17:22 |
*** Rockyg has quit IRC | 17:26 | |
*** Apoorva_ has joined #openstack-operators | 17:30 | |
*** rebase has quit IRC | 17:31 | |
*** rebase has joined #openstack-operators | 17:32 | |
*** makowals_ has joined #openstack-operators | 17:33 | |
*** Apoorva has quit IRC | 17:33 | |
*** makowals has quit IRC | 17:34 | |
*** slaweq has joined #openstack-operators | 17:37 | |
*** slaweq has quit IRC | 17:40 | |
*** pcaruana has quit IRC | 17:40 | |
*** electrofelix has quit IRC | 18:01 | |
*** aojea has quit IRC | 18:01 | |
*** aojea has joined #openstack-operators | 18:04 | |
*** dminer has joined #openstack-operators | 18:08 | |
*** slaweq has joined #openstack-operators | 18:11 | |
*** slaweq has quit IRC | 18:17 | |
*** slaweq has joined #openstack-operators | 18:18 | |
*** slaweq has quit IRC | 18:22 | |
*** aojea has quit IRC | 18:24 | |
*** aojea has joined #openstack-operators | 18:25 | |
yankcrime | lol @ erhudy | 18:25 |
yankcrime | that's basically us right now | 18:26 |
*** gyee has joined #openstack-operators | 18:26 | |
yankcrime | moving from one super EOL release to only a vaguely EOL release | 18:26 |
yankcrime | with a short-term goal to get to mitaka, which is just about to go EOL | 18:26 |
yankcrime | fml basically | 18:26 |
erhudy | liberty's been fine for us, we'll jump them up to mitaka and they'll stick around for a few more years | 18:27 |
*** aojea has quit IRC | 18:29 | |
yankcrime | yeah, it's a desire to layer on some of the newer features such as magnum that's prompted us to get a bit more aggressive with upgrades | 18:30 |
yankcrime | sticking stuff in containers has really helped | 18:30 |
*** Apoorva_ has quit IRC | 18:37 | |
*** Apoorva has joined #openstack-operators | 18:38 | |
*** fragatina has quit IRC | 18:46 | |
*** fragatina has joined #openstack-operators | 18:46 | |
*** marst_ has joined #openstack-operators | 18:49 | |
*** shasha___ has quit IRC | 18:49 | |
klindgren_ | thats what we are going through right now | 18:52 |
*** usertm has quit IRC | 18:52 | |
klindgren_ | moving our liberty install into containers | 18:52 |
klindgren_ | we also have a number of patches around cellsv1 that makes upgrading "interesting" | 18:52 |
klindgren_ | especially since nova started moving everything to objects and breaking cells v1 code all over the place | 18:53 |
*** VW_ has joined #openstack-operators | 18:57 | |
*** aojea has joined #openstack-operators | 18:58 | |
*** VW_ has quit IRC | 18:59 | |
*** VW_ has joined #openstack-operators | 18:59 | |
*** VW has quit IRC | 19:01 | |
*** VW_ has quit IRC | 19:03 | |
*** VW has joined #openstack-operators | 19:03 | |
*** fragatina has quit IRC | 19:04 | |
*** fragatina has joined #openstack-operators | 19:04 | |
*** racedo has quit IRC | 19:18 | |
*** racedo has joined #openstack-operators | 19:20 | |
*** marst_ has quit IRC | 19:22 | |
*** slaweq has joined #openstack-operators | 19:23 | |
*** robksawyer has joined #openstack-operators | 19:23 | |
*** slaweq has quit IRC | 19:28 | |
*** racedo has quit IRC | 19:29 | |
*** manheim has joined #openstack-operators | 19:29 | |
*** robksawyer has quit IRC | 19:29 | |
*** arnewiebalck_ has joined #openstack-operators | 19:30 | |
*** arnewiebalck_ has quit IRC | 19:30 | |
*** rebase has quit IRC | 19:31 | |
*** arnewiebalck_ has joined #openstack-operators | 19:32 | |
*** rebase has joined #openstack-operators | 19:32 | |
*** rmart04 has quit IRC | 19:33 | |
*** shewless_ has joined #openstack-operators | 19:55 | |
shewless_ | Hello. Does anyone have any good reference material for integrating ceph with openstack. Specifically for using ceph for ephemeral storage (what kind of iops is required) | 19:56 |
*** Apoorva has quit IRC | 19:59 | |
*** dminer has quit IRC | 20:00 | |
*** fandi has quit IRC | 20:04 | |
*** aojea has quit IRC | 20:08 | |
*** rockyg has joined #openstack-operators | 20:16 | |
*** aojea has joined #openstack-operators | 20:17 | |
*** Apoorva has joined #openstack-operators | 20:17 | |
*** VW has quit IRC | 20:20 | |
*** VW has joined #openstack-operators | 20:21 | |
*** dminer has joined #openstack-operators | 20:23 | |
*** dbecker has quit IRC | 20:25 | |
shewless_ | I'm not looking for "how" to integrate ceph with openstack. I'm moreso looking for the performance requirements of the ceph cluster | 20:28 |
*** rockyg has quit IRC | 20:32 | |
zerick | shewless_: I guess Red Hat docs would point on that direction | 20:38 |
*** rockyg has joined #openstack-operators | 20:38 | |
shewless_ | zerick: thanks.. I'm not seeing anything on that subject | 20:39 |
*** arnewiebalck_ has quit IRC | 20:41 | |
*** manheim has quit IRC | 20:48 | |
*** Fin1te has joined #openstack-operators | 21:08 | |
*** haplo37 has quit IRC | 21:09 | |
*** slaweq has joined #openstack-operators | 21:16 | |
*** Caterpillar has quit IRC | 21:16 | |
*** haplo37 has joined #openstack-operators | 21:18 | |
*** dminer has quit IRC | 21:22 | |
*** slaweq has quit IRC | 21:30 | |
*** Apoorva_ has joined #openstack-operators | 21:34 | |
*** Apoorva_ has quit IRC | 21:36 | |
*** Apoorva_ has joined #openstack-operators | 21:36 | |
*** Apoorva has quit IRC | 21:36 | |
zerick | shewless_: maybe this? http://www3.intel.com/content/dam/www/public/us/en/documents/reference-architectures/redhat-cloud-blocks-solution-brief.pdf | 21:43 |
zerick | and this https://access.redhat.com/sites/default/files/attachments/osp10-rhcs2-v1.pdf | 21:43 |
*** rockyg has quit IRC | 21:47 | |
*** chlong has quit IRC | 21:48 | |
*** Fin1te has quit IRC | 21:57 | |
*** manheim has joined #openstack-operators | 21:59 | |
*** Apoorva_ has quit IRC | 22:02 | |
*** manheim has quit IRC | 22:03 | |
*** Apoorva has joined #openstack-operators | 22:03 | |
*** shasha___ has joined #openstack-operators | 22:09 | |
*** aojea has quit IRC | 22:10 | |
*** jamesdenton has quit IRC | 22:12 | |
*** manheim has joined #openstack-operators | 22:25 | |
*** slaweq has joined #openstack-operators | 22:31 | |
*** VW_ has joined #openstack-operators | 22:32 | |
*** slaweq has quit IRC | 22:35 | |
*** VW has quit IRC | 22:36 | |
*** VW_ has quit IRC | 22:36 | |
*** shasha___ has quit IRC | 22:47 | |
*** Apoorva_ has joined #openstack-operators | 23:10 | |
*** Apoorva_ has quit IRC | 23:11 | |
*** Apoorva has quit IRC | 23:11 | |
*** Apoorva has joined #openstack-operators | 23:11 | |
*** manheim has quit IRC | 23:21 | |
*** shasha___ has joined #openstack-operators | 23:22 | |
*** shasha___ has quit IRC | 23:27 | |
*** chyka has quit IRC | 23:32 | |
*** shasha___ has joined #openstack-operators | 23:36 | |
*** shasha___ has quit IRC | 23:43 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!