| *** jaewook_oh has joined #openstack-containers | 01:30 | |
| *** hongbin has joined #openstack-containers | 02:24 | |
| *** ramishra has joined #openstack-containers | 03:18 | |
| *** chhagarw has joined #openstack-containers | 03:50 | |
| *** udesale has joined #openstack-containers | 03:57 | |
| *** udesale has quit IRC | 04:07 | |
| *** udesale has joined #openstack-containers | 04:20 | |
| *** hongbin has quit IRC | 04:28 | |
| *** pcaruana has quit IRC | 06:37 | |
| *** Bhujay has joined #openstack-containers | 06:47 | |
| *** Bhujay has quit IRC | 06:49 | |
| *** pcaruana has joined #openstack-containers | 07:02 | |
| *** rcernin has quit IRC | 07:04 | |
| *** rcernin has joined #openstack-containers | 07:08 | |
| *** rcernin has quit IRC | 07:21 | |
| *** mattgo has joined #openstack-containers | 07:36 | |
| *** ttsiouts has joined #openstack-containers | 08:04 | |
| *** ttsiouts has quit IRC | 08:06 | |
| *** kaiokmo has quit IRC | 08:07 | |
| *** kaiokmo has joined #openstack-containers | 08:07 | |
| *** ttsiouts has joined #openstack-containers | 08:17 | |
| *** ttsiouts has quit IRC | 08:22 | |
| *** ttsiouts has joined #openstack-containers | 08:24 | |
| *** Bhujay has joined #openstack-containers | 08:36 | |
| *** Bhujay has quit IRC | 08:37 | |
| *** Bhujay has joined #openstack-containers | 08:37 | |
| brtknr | what do people in this channel use for their magnum dev environment? | 08:41 |
|---|---|---|
| brtknr | i used to have a devstack on baremetal that worked okay... however, i have not had much luck getting devstack with magnum enabled up (it works fine without magnum) | 08:42 |
| brtknr | ... inside a virtual machine that is | 08:44 |
| mattgo | brtknr, Here we are using DevStack in a VM, Magnum enabled, Queens. Works fine, to some extent. | 08:53 |
| mattgo | brtknr, I found it hard to find the correct lines to add to the local.conf file (documentation is hard to find..), but now it works fine. We are regulary spawning new instances of DevStack, it's stable. For the record, we have those additional lines in local.conf : | 08:56 |
| mattgo | enable_plugin heat https://git.openstack.org/openstack/heat $TARGET_BRANCH | 08:56 |
| mattgo | enable_plugin magnum https://git.openstack.org/openstack/magnum $TARGET_BRANCH | 08:56 |
| mattgo | enable_plugin magnum-ui https://github.com/openstack/magnum-ui $TARGET_BRANCH | 08:56 |
| mattgo | enable_plugin python-magnumclient https://github.com/openstack/python-magnumclient $TARGET_BRANCH | 08:56 |
| *** ttsiouts has quit IRC | 09:11 | |
| *** ttsiouts has joined #openstack-containers | 09:12 | |
| *** ttsiouts has quit IRC | 09:16 | |
| *** ttsiouts has joined #openstack-containers | 09:17 | |
| brtknr | mattgo: thanks! what does your full local.conf look like? | 09:21 |
| brtknr | did you also need to configure fixed subnet? | 09:21 |
| *** salmankhan has joined #openstack-containers | 09:22 | |
| mattgo | brtknr, The full local.conf is not very much complex : http://paste.openstack.org/show/732146/ | 09:28 |
| mattgo | brtknr, You could fix a subnet (we actually do fix the public network range), but that shouldn't be necessary. | 09:28 |
| mattgo | brtknr, Also, what you don't see here, we have are using a Heat Stack to deploy DevStack + Configure Keytsone to Keystone federation + install additional softwares on the VM. We use a DevStack ID to differenciate our multiple DevStack instances. I don't think this is relevant, just want to explain why you see $DEVSTACK_ID in the local.conf | 09:30 |
| *** lpetrut has joined #openstack-containers | 09:35 | |
| *** Bhujay has quit IRC | 09:47 | |
| *** Bhujay has joined #openstack-containers | 09:54 | |
| *** ramishra has quit IRC | 09:55 | |
| *** Bhujay has quit IRC | 09:55 | |
| *** Bhujay has joined #openstack-containers | 09:56 | |
| *** ttsiouts has quit IRC | 10:19 | |
| *** ramishra has joined #openstack-containers | 10:23 | |
| *** udesale has quit IRC | 11:24 | |
| brtknr | mattgo: nice! i was wondering that the $DEVSTACK_ID was | 11:25 |
| brtknr | i currently have 2 ./stack.sh command running, one on bare metal and another on VM.... will wait and see if it has finally worked | 11:27 |
| brtknr | unfortunately, the VM is running in an openstack deployment deployed using default config.... so the parent subnet is 10.0.0.0/24, same as devstack default subnet | 11:28 |
| brtknr | so it keeps having issues... | 11:28 |
| brtknr | and not always the same... sometimes its related to python version, other times, related to lvm... and mostly it can't reach glance-api on 127.0.0.1/image | 11:29 |
| *** ttsiouts has joined #openstack-containers | 11:40 | |
| *** kaiokmo has quit IRC | 11:54 | |
| *** kaiokmo has joined #openstack-containers | 11:54 | |
| *** canori01 has joined #openstack-containers | 11:57 | |
| *** zul has joined #openstack-containers | 12:21 | |
| *** ttsiouts has quit IRC | 12:25 | |
| *** ttsiouts has joined #openstack-containers | 12:35 | |
| mattgo | brtknr, I don't exactly understand why the devstack default subnet being the same as the underlying openstack subnet is a problem, what are you trying to achieve ? To spawn DevStack with Magnum and play with it, the underlying subnet shouldn't matter in my understanding. Now, if your goal is to access your instances deployed within DevStack from the external world, then you'll have some additional work (but that's unrelated to Magnum) | 12:53 |
| *** jaewook_oh has quit IRC | 12:54 | |
| mattgo | brtknr, Also regarding python version, etc, I just can tell you what we are doing before ./stack.sh : From Ubuntu 16.04, enable nested virtualization, add stack user with sudo perm, clone DevStack (sudo -H -u stack git clone https://git.openstack.org/openstack-dev/devstack -b stable/queens), and then run ./stack.sh | 12:56 |
| mattgo | brtknr, DevStack is a sensible beast, finding the right config was tricky and sometimes I didn't exactly know why this or this option would make things work, or brake things... As I said, good documentation is hard to find. | 12:58 |
| *** dave-mccowan has joined #openstack-containers | 13:43 | |
| *** robertomls has joined #openstack-containers | 13:59 | |
| robertomls | are we having any known problems with the zuul magnum-functional-api gate? I uploaded a patch set on the weekend and it failed there but I'm not sure what I did.. | 14:03 |
| robertomls | this is the one: https://review.openstack.org/#/c/598142/ | 14:03 |
| robertomls | I saw other builds failing at that gate too so just checking if it's something known already | 14:04 |
| *** Bhujay has quit IRC | 14:06 | |
| *** hongbin has joined #openstack-containers | 14:08 | |
| *** munimeha1 has joined #openstack-containers | 14:11 | |
| tobias-urdin | strigazi: did you see my message some days ago about releasing 7.0.2 for rocky? | 14:34 |
| *** Bhujay has joined #openstack-containers | 15:02 | |
| brtknr | mattgo: turns out the problem was actually due to default security policy which meant that i couldnt ping/ssh my nodes by default and had to add these policies myself | 15:03 |
| brtknr | mattgo: do you know why docker cannot pull images from its various sources? i can ping 1.1.1.1 however | 15:04 |
| brtknr | mattgo: my resolv.conf is using my virtual router as dns resolver | 15:06 |
| brtknr | how do i make my router resolve dns correctly? | 15:06 |
| brtknr | i know that i can add dns argument to my magnum cluster template | 15:07 |
| mattgo | brtknr, Glad you could resolve some of your issues. Are you working in a DevStack VM deployed in OpenStack or on BM ? where are you trying to do docker pull: DevStack itself, or an instance within DevStack ? | 15:10 |
| strigazi | brtknr: I'm using a physical machine, works fine for me. | 15:13 |
| strigazi | tobias-urdin: no, but I saw this one. We need two more patches and I'll release ti | 15:13 |
| tobias-urdin | cool, thanks! | 15:14 |
| *** Bhujay has quit IRC | 15:14 | |
| brtknr | mattgo: strigazi: I'm using DevStack VM deploying in Openstack. BM setup didnt quite work out, it was too slow | 15:17 |
| brtknr | mattgo: strigazi: I'm using DevStack VM deploying in Openstack. BM setup didnt quite work out, it ran into errors, will try again shortly | 15:17 |
| strigazi | brtknr: it takes 30mins for me, 4-core i7, 16GB ram, 500GB ssd | 15:18 |
| brtknr | strigazi: to setup DevStack or to deploy k8s cluster? | 15:19 |
| strigazi | brtknr: to setup devstack | 15:19 |
| brtknr | strigazi: Ah, I was talking about deploying k8s via magnum | 15:20 |
| strigazi | brtknr: with bm nodes? | 15:20 |
| mattgo | To deploy DevStack VM in OpenStack, it also takes 30 minutes for me. Then, deploy a k8s cluster in this DevStack VM is rather slow, I think about 20 minutes. Interestingly, the more k8s cluster you deploy, the slower it gets to deploy new ones :/ | 15:21 |
| *** ttsiouts has quit IRC | 15:21 | |
| mattgo | but if you can, I would suggest to deploy DevStack on BM directly. It should be faster, and you avoid one level of abstraction, the intermediate OpenStack, which makes things way more simple IMO. | 15:22 |
| *** ramishra has quit IRC | 15:31 | |
| brtknr | Yeah I will try that | 15:32 |
| *** Bhujay has joined #openstack-containers | 15:32 | |
| brtknr | trying to use magnum inside a vm is sooo slow | 15:32 |
| strigazi | do you have you meed nested kvm enabled? | 15:32 |
| *** munimeha1 has quit IRC | 15:33 | |
| *** jmlowe has quit IRC | 15:37 | |
| *** ttsiouts has joined #openstack-containers | 15:40 | |
| brtknr | So ive been told | 15:40 |
| brtknr | is there a way to check this? | 15:40 |
| strigazi | cat /etc/cpuinfo | grep vmx | 15:40 |
| brtknr | yields nil | 15:43 |
| brtknr | hmm | 15:43 |
| brtknr | guess that means it isnt | 15:43 |
| strigazi | cat /etc/cpuinfo | grep svm | 15:44 |
| strigazi | these are the only two modules | 15:44 |
| strigazi | but usually it is vmx | 15:44 |
| brtknr | dont you mean `cat /proc/cpuinfo | grep vmx|svm` | 15:44 |
| brtknr | no neither are loaded | 15:45 |
| strigazi | so , it is not | 15:46 |
| *** openstackgerrit has quit IRC | 15:47 | |
| *** openstackgerrit has joined #openstack-containers | 15:48 | |
| brtknr | does it make a huge difference? | 15:49 |
| strigazi | it does. the overhead is minimal | 15:50 |
| strigazi | the only thing that won't change is if the hypervisor has N disks and N+M vms doing IO | 15:50 |
| *** robertomls has quit IRC | 15:52 | |
| *** munimeha1 has joined #openstack-containers | 15:53 | |
| *** ttsiouts has quit IRC | 16:10 | |
| *** ttsiouts has joined #openstack-containers | 16:11 | |
| *** ttsiouts has quit IRC | 16:15 | |
| *** jmlowe has joined #openstack-containers | 16:17 | |
| *** lpetrut has quit IRC | 16:37 | |
| brtknr | strigazi: just had this error while trying to deploy devstack on BM: http://paste.openstack.org/show/732160/ | 16:40 |
| brtknr | strigazi: after about 28 mins :( | 16:40 |
| brtknr | s/:(/:'( | 16:41 |
| brtknr | looks like this was when this was triying to run: `/usr/bin/magnum-db-manage upgrade` | 16:44 |
| *** Bhujay has quit IRC | 16:46 | |
| brtknr | strigazi: this is on Centos 7.4 | 16:47 |
| *** lpetrut has joined #openstack-containers | 17:04 | |
| *** salmankhan has quit IRC | 17:10 | |
| strigazi | brtknr: sorry, I was commuting | 17:11 |
| *** itlinux has joined #openstack-containers | 17:12 | |
| strigazi | brtknr: this error is not related to vm vs bm, for sure. | 17:12 |
| strigazi | brtknr: Can you share your local.conf? | 17:12 |
| strigazi | brtknr: can try to run magnum-db-manage upgrade again? | 17:13 |
| *** chhagarw has quit IRC | 18:00 | |
| *** lpetrut has quit IRC | 18:43 | |
| *** jmlowe has quit IRC | 18:44 | |
| *** hongbin has quit IRC | 18:57 | |
| *** jmlowe has joined #openstack-containers | 19:19 | |
| *** mattgo has quit IRC | 19:36 | |
| brtknr | strigazi: I tried doing that, same error | 19:56 |
| brtknr | I changed my max_connections value from 1024 to 10240 | 19:57 |
| brtknr | under devstack/lib/databases/mysql | 20:05 |
| *** hongbin has joined #openstack-containers | 20:30 | |
| *** hongbin has quit IRC | 20:33 | |
| *** hongbin has joined #openstack-containers | 20:36 | |
| brtknr | is there something about deploying devstack on centos thats buggy? | 21:30 |
| brtknr | i understand that ubuntu is whats usually recommended | 21:31 |
| brtknr | I just noticed this: | 21:44 |
| brtknr | MariaDB [(none)]> show variables like "max_connections"; | 21:44 |
| brtknr | +-----------------+-------+ | 21:44 |
| brtknr | | Variable_name | Value | | 21:44 |
| brtknr | +-----------------+-------+ | 21:44 |
| brtknr | | max_connections | 214 | | 21:44 |
| brtknr | +-----------------+-------+ | 21:44 |
| brtknr | 1 row in set (0.00 sec) | 21:44 |
| *** openstackgerrit has quit IRC | 21:56 | |
| *** openstackgerrit has joined #openstack-containers | 21:57 | |
| *** openstackgerrit has quit IRC | 21:58 | |
| *** openstackgerrit has joined #openstack-containers | 22:00 | |
| brtknr | im about to try setting open_files_limit=1024 and trying to bring devstack up again! will keep it running and check tomorrow morning/post result | 22:06 |
| mordred | strigazi: I can't remember if I got your attention the other day - but did I pointyou at the broken magnum jobs in the openstacksdk gate? | 22:12 |
| mordred | strigazi: http://logs.openstack.org/18/608318/5/check/openstacksdk-functional-devstack-magnum/95e905a/ is an example | 22:12 |
| mordred | (or anyone else who knows things about devstack+magnum) | 22:12 |
| *** openstackgerrit has quit IRC | 22:12 | |
| mordred | it's non-voting, so it's not killing us, but I'm worried that we're not verifying magnum support now | 22:13 |
| *** openstackgerrit has joined #openstack-containers | 22:14 | |
| openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/magnum master: [DNM] Testing amqp gate changes https://review.openstack.org/610740 | 22:23 |
| *** hongbin has quit IRC | 22:32 | |
| *** hongbin has joined #openstack-containers | 22:34 | |
| *** munimeha1 has quit IRC | 22:34 | |
| *** rcernin has joined #openstack-containers | 22:41 | |
| *** openstackgerrit has quit IRC | 22:43 | |
| *** itlinux has quit IRC | 22:54 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!