| *** vipul-away is now known as vipul | 00:00 | |
| *** krotscheck has joined #tripleo | 00:09 | |
| lifeless | SpamapS: all mails sent :) | 00:10 |
|---|---|---|
| SpamapS | lifeless: sweet | 00:14 |
| SpamapS | lifeless: left a comment on one of your rebuild reviews | 00:14 |
| SpamapS | lifeless: want me to take a crack at tests/commentary for 60285 ? | 00:15 |
| SpamapS | lifeless: also remember that Heat actually already does have rebuild support. | 00:16 |
| lifeless | SpamapS: doesn't have the 'ephemeral' bit set for it | 00:19 |
| lifeless | SpamapS: does it? | 00:19 |
| lifeless | it's an active opt-in behaviour | 00:19 |
| SpamapS | lifeless: oh right we'd have to add that bit. | 00:21 |
| SpamapS | lifeless: very very tiny change. | 00:21 |
| lifeless | yes | 00:22 |
| SpamapS | basically just a property we set to true. | 00:22 |
| lifeless | yes | 00:22 |
| lifeless | SpamapS: and yes, crack on on the patch | 00:22 |
| *** cd-undercloud has joined #tripleo | 00:22 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 00:22 |
| *** cd-undercloud has quit IRC | 00:22 | |
| *** AaronGr is now known as AaronGr_afk | 00:28 | |
| *** vipul is now known as vipul-away | 00:39 | |
| openstackgerrit | Clint "SpamapS" Byrum proposed a change to openstack/tripleo-incubator: Make refresh-env more resilient for users https://review.openstack.org/61617 | 00:39 |
| *** sdake_ has quit IRC | 00:49 | |
| *** morazi has quit IRC | 00:59 | |
| *** vipul-away is now known as vipul | 01:10 | |
| *** jcooley_ has quit IRC | 01:12 | |
| *** jergerber has quit IRC | 01:13 | |
| *** cd-undercloud has joined #tripleo | 01:15 | |
| cd-undercloud | ************** overcloud complete status=42 ************ | 01:15 |
| *** cd-undercloud has quit IRC | 01:15 | |
| *** jcooley_ has joined #tripleo | 01:33 | |
| *** nosnos has joined #tripleo | 01:35 | |
| *** tserong has quit IRC | 01:42 | |
| *** csd has quit IRC | 01:48 | |
| *** csd has joined #tripleo | 01:50 | |
| *** rongze has joined #tripleo | 01:52 | |
| *** krotscheck has quit IRC | 01:58 | |
| *** csd has quit IRC | 02:05 | |
| *** rcarrillocruz has joined #tripleo | 02:06 | |
| *** tserong has joined #tripleo | 02:07 | |
| *** cd-undercloud has joined #tripleo | 02:09 | |
| cd-undercloud | ************** overcloud complete status=1 ************ | 02:09 |
| *** cd-undercloud has quit IRC | 02:09 | |
| *** rcarrillocruz has quit IRC | 02:11 | |
| *** arata has joined #tripleo | 02:19 | |
| *** jcooley_ has quit IRC | 02:24 | |
| *** rongze has quit IRC | 02:28 | |
| *** rongze has joined #tripleo | 02:29 | |
| *** vkozhukalov has joined #tripleo | 02:38 | |
| *** krotscheck has joined #tripleo | 02:48 | |
| *** krotscheck has quit IRC | 02:49 | |
| *** julim has quit IRC | 02:57 | |
| *** cd-undercloud has joined #tripleo | 03:03 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 03:03 |
| *** cd-undercloud has quit IRC | 03:03 | |
| *** rpodolyaka has quit IRC | 03:12 | |
| *** rpodolyaka has joined #tripleo | 03:13 | |
| *** Haomeng has quit IRC | 03:20 | |
| *** Haomeng has joined #tripleo | 03:20 | |
| *** jcooley_ has joined #tripleo | 03:20 | |
| openstackgerrit | A change was merged to openstack/diskimage-builder: Add -U to pip install command in tox.ini https://review.openstack.org/61502 | 03:24 |
| *** jcooley_ has quit IRC | 03:26 | |
| *** rongze_ has joined #tripleo | 03:33 | |
| *** rongze has quit IRC | 03:36 | |
| *** cd-undercloud has joined #tripleo | 03:55 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 03:55 |
| *** cd-undercloud has quit IRC | 03:55 | |
| *** sdake_ has joined #tripleo | 04:05 | |
| *** marun has joined #tripleo | 04:06 | |
| *** rushiagr has joined #tripleo | 04:06 | |
| *** AaronGr_afk is now known as AaronGr | 04:08 | |
| *** morazi has joined #tripleo | 04:09 | |
| *** AaronGr is now known as AaronGr_Zzz | 04:09 | |
| *** ccrouch has joined #tripleo | 04:09 | |
| *** rushiagr has quit IRC | 04:17 | |
| *** ohadlevy has quit IRC | 04:18 | |
| *** ohadlevy has joined #tripleo | 04:19 | |
| *** ohadlevy is now known as Guest85245 | 04:19 | |
| *** jergerber has joined #tripleo | 04:25 | |
| *** Guest85245 has quit IRC | 04:37 | |
| *** jergerber has quit IRC | 04:39 | |
| *** ohadlevy_ has joined #tripleo | 04:45 | |
| *** cd-undercloud has joined #tripleo | 04:49 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 04:49 |
| *** cd-undercloud has quit IRC | 04:49 | |
| *** rongze_ has quit IRC | 04:52 | |
| *** rongze has joined #tripleo | 05:22 | |
| *** rongze has quit IRC | 05:27 | |
| *** rpodolyaka1 has joined #tripleo | 05:32 | |
| *** sdake_ has quit IRC | 05:32 | |
| rpodolyaka1 | morning tripleo | 05:32 |
| lifeless | morning! | 05:33 |
| rpodolyaka1 | lifeless: hey! I've a got a small update on rebuild. Our baremetal 'specialized' rebuild works again (without preserving ephemeral right now) | 05:35 |
| lifeless | cool | 05:36 |
| rpodolyaka1 | lifeless: haven't tested preserving yet, because I didn't manage to boot a BM node with ephemeral >=0 mb | 05:36 |
| rpodolyaka1 | lifeless: the scheduler couldn't find a node | 05:36 |
| rpodolyaka1 | lifeless: going to debug this further today. But all we need is a different flavor with ephemeral_mb >=0, right? | 05:37 |
| rpodolyaka1 | lifeless: our registered nodes should me modified too? haven't looked at how they are registered yet | 05:38 |
| rpodolyaka1 | s/our/or | 05:38 |
| lifeless | rpodolyaka1: our automation scripts haven't been updated | 05:39 |
| lifeless | but the cd-undercloud has ephemeral defined | 05:39 |
| lifeless | rpodolyaka1: what you need to do is subtract X space from the root and add it to ephemeral in the flavor definition | 05:39 |
| lifeless | they should sum to the prior value | 05:39 |
| *** cd-undercloud has joined #tripleo | 05:40 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 05:40 |
| *** cd-undercloud has quit IRC | 05:40 | |
| rpodolyaka1 | lifeless: I believe, this is exactly what I was doing. Ok, going to debug this today then | 05:41 |
| lifeless | rpodolyaka1: so that deploy ^ that just worked, has ephemeral volume | 05:41 |
| rpodolyaka1 | hmm... | 05:42 |
| lifeless | running trunk | 05:42 |
| lifeless | trunk | 05:42 |
| *** tzumainn has quit IRC | 05:43 | |
| rpodolyaka1 | oh, and btw, the fix for https://bugs.launchpad.net/neutron/+bug/1254246 has been merged | 05:43 |
| uvirtbot | Launchpad bug 1254246 in tripleo "somehow getting duplicate openvswitch agents for the same host" [Critical,In progress] | 05:43 |
| lifeless | cool | 05:43 |
| rpodolyaka1 | though I think we have a new neutron issue with "vif binding failed" now | 05:44 |
| lifeless | thats a race condition with the agent | 05:44 |
| rpodolyaka1 | saw it a few times yesterday | 05:44 |
| lifeless | SpamapS has a patch up to workaround it | 05:44 |
| rpodolyaka1 | cool! | 05:44 |
| lifeless | sda 8:0 0 1.8T 0 disk | 05:45 |
| lifeless | ├─sda1 8:1 0 1000G 0 part /mnt | 05:45 |
| lifeless | ├─sda2 8:2 0 7.9M 0 part | 05:45 |
| lifeless | └─sda3 8:3 0 500G 0 part / | 05:45 |
| lifeless | ^ lsblk on the deployed overcloud node | 05:45 |
| *** rushiagr has joined #tripleo | 05:45 | |
| rpodolyaka1 | nice! let's make it preservable ;) | 05:46 |
| lifeless | this is the flavor definition | 05:50 |
| lifeless | http://paste.ubuntu.com/6559674/ | 05:50 |
| rpodolyaka1 | thanks, will compare it to mine | 05:51 |
| StevenK | lifeless: Didn't know about lsblk, that looks handy | 06:05 |
| *** jcooley_ has joined #tripleo | 06:06 | |
| *** rongze has joined #tripleo | 06:20 | |
| *** vipul has quit IRC | 06:22 | |
| *** vipul has joined #tripleo | 06:22 | |
| *** jcooley_ has quit IRC | 06:30 | |
| *** jcooley_ has joined #tripleo | 06:30 | |
| *** jcooley_ has quit IRC | 06:31 | |
| *** jcooley_ has joined #tripleo | 06:31 | |
| *** cd-undercloud has joined #tripleo | 06:32 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 06:32 |
| *** cd-undercloud has quit IRC | 06:32 | |
| *** lifeless has quit IRC | 06:48 | |
| *** vkozhukalov has quit IRC | 06:51 | |
| *** lifeless has joined #tripleo | 07:15 | |
| *** ChanServ sets mode: +v lifeless | 07:15 | |
| *** jcooley_ has quit IRC | 07:18 | |
| *** jcooley_ has joined #tripleo | 07:22 | |
| *** cd-undercloud has joined #tripleo | 07:26 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 07:26 |
| *** cd-undercloud has quit IRC | 07:26 | |
| *** jcoufal has joined #tripleo | 07:27 | |
| *** jprovazn has joined #tripleo | 07:39 | |
| *** boris-42 has quit IRC | 07:43 | |
| *** ohadlevy_ is now known as ohadlevy | 07:45 | |
| *** ohadlevy has joined #tripleo | 07:46 | |
| *** jcooley_ has quit IRC | 07:46 | |
| *** rcarrillocruz has joined #tripleo | 07:49 | |
| *** jcooley_ has joined #tripleo | 07:50 | |
| *** rpodolyaka1 has quit IRC | 07:56 | |
| *** ifarkas has joined #tripleo | 07:56 | |
| *** jcooley_ has quit IRC | 07:58 | |
| *** rongze has quit IRC | 08:02 | |
| *** rongze has joined #tripleo | 08:06 | |
| *** rongze has quit IRC | 08:11 | |
| lifeless | jprovazn: hi | 08:14 |
| lifeless | jprovazn: can we chat about the haproxy patches a little? | 08:15 |
| jprovazn | lifeless: Hi, sure, I'm now reading your replies | 08:19 |
| jprovazn | lifeless: about haproxy and keepalive | 08:20 |
| jprovazn | both elements need to know virtual ip | 08:20 |
| *** cd-undercloud has joined #tripleo | 08:20 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 08:20 |
| *** cd-undercloud has quit IRC | 08:20 | |
| jprovazn | lifeless: the mistake was to use haproxy.virtual_ip | 08:21 |
| jprovazn | lifeless: the config in template can be put outside of haproxy | 08:21 |
| jprovazn | I used haproxy.virtual_ip in keepalived just because I already had it in template, but you are right that it's misleading | 08:24 |
| lifeless | jprovazn: so I guess I want to talk at a meta level | 08:24 |
| lifeless | like - why a virtual IP for haproxy? Whats the topology you're implementing (and how much variation do we want/need to support today?) | 08:25 |
| lifeless | I don't want to create a boil-the-oceans scenario for you | 08:25 |
| lifeless | but equally I don't want to head down the wrong path early on! | 08:25 |
| jprovazn | ok, so virtual ip is needed to make haproxy not being single point of failure | 08:27 |
| jprovazn | so the topology I wanted to use | 08:28 |
| *** vkozhukalov has joined #tripleo | 08:28 | |
| lifeless | jprovazn: oh, so previously I've just run N in active-active | 08:28 |
| lifeless | with DNS round-robin | 08:28 |
| lifeless | do we have clients that don't handle that, or does virtual IP offer faster failover ? | 08:28 |
| jprovazn | just run haproxy+keepalived on all controller nodes, virtual ip would be used to point to active haproxy | 08:28 |
| lifeless | if there is keepalived everywhere, and a virtual ip that can move to any node, whats haproxy doing in that setup ? | 08:29 |
| jprovazn | lifeless: points to the specific running service | 08:30 |
| jprovazn | virtual ip is used on node level, haproxy on service level (at least this is my understanding) | 08:30 |
| lifeless | so our services fall into two categories | 08:31 |
| lifeless | mmm three | 08:31 |
| lifeless | active-active | 08:31 |
| lifeless | active-passive | 08:32 |
| lifeless | and self-managing | 08:32 |
| lifeless | so nova-api is active-active | 08:32 |
| lifeless | run as many as we want | 08:32 |
| lifeless | heat-engine today is active-passive - run one, start a new one when that fails | 08:32 |
| lifeless | self managing are things like extradb | 08:32 |
| lifeless | we probably need to analyse these separately | 08:33 |
| lifeless | active-active, if the virtual IP has moved, the service on the machine can just answer directly. Done - listen to 0.0.0.0 and it will Just Work. | 08:33 |
| lifeless | self-managing we don't want keepalived fiddling with at all. | 08:34 |
| jprovazn | mmt | 08:34 |
| jprovazn | "if the virtual ip has moved" - keepalived does this | 08:34 |
| lifeless | active-passive, if the virtual IP has moved, wouldn't that also mean the origin machine is either being taken offline (manual move) or has failed (automatic move) ? | 08:34 |
| lifeless | sure, I'm rolling with using vip, whether it's needed or not is perhaps a separated iscussion | 08:35 |
| lifeless | you've done work to get it working - cool. | 08:35 |
| jprovazn | lifeless: well, if VIP has moved it just means that the original one stopped responding | 08:35 |
| lifeless | righto | 08:35 |
| jprovazn | well | 08:36 |
| jprovazn | https://etherpad.openstack.org/p/tripleo-icehouse-ha-production-configuration | 08:36 |
| lifeless | so I'm just failing to understand when haproxy would send traffic to a difference machnie | 08:36 |
| jprovazn | *api instances | 08:36 |
| jprovazn | *virtual IP - keepalived + haproxy | 08:36 |
| jprovazn | IOW ,my understanding was | 08:36 |
| lifeless | and if haproxy isn't sending traffic to a different machine, what's it for :0 | 08:36 |
| jprovazn | ok, so I can imagine 2 cases: | 08:37 |
| *** akrivoka has joined #tripleo | 08:37 | |
| jprovazn | situation - I have 2 controller nodes, both run API services | 08:37 |
| jprovazn | 1) on both I have openstack services configured to point to my virtual ip | 08:38 |
| jprovazn | virtual ip points to one of these controller nodes | 08:38 |
| jprovazn | if nova-api on node1 fails, then haproxy on controller node where virtual ip points to just forwards nova requests to the second node | 08:39 |
| jprovazn | but other services forwards as before | 08:39 |
| jprovazn | 2) if my controller node where VIP points to blows out | 08:40 |
| lifeless | ok | 08:40 |
| jprovazn | the seond node takesover VIP and requests go through haproxy on this second machine | 08:40 |
| lifeless | lets roll with this | 08:40 |
| lifeless | the implication is - let me check I have it right: | 08:40 |
| lifeless | - haproxy runs on set of VIP participating nodes [e.g. all, or maybe we build dedicated image] | 08:41 |
| jprovazn | yes | 08:41 |
| lifeless | - each haproxy needs connectivity to all backend apis | 08:41 |
| lifeless | so - here's what I propose | 08:42 |
| jprovazn | yes | 08:42 |
| lifeless | don't fiddle with the binding port for any of the services. | 08:42 |
| lifeless | at least for openstack services other than the keystone discovery port. | 08:42 |
| lifeless | everything else has the port encoded in the end point. | 08:42 |
| lifeless | So lets make the super repeatable and predictable: | 08:43 |
| lifeless | - always install haproxy | 08:43 |
| lifeless | oh | 08:43 |
| lifeless | hmm hang on a second | 08:43 |
| lifeless | whats our SSL strategy here? | 08:43 |
| jprovazn | I don't think we have SSL strategy so far | 08:44 |
| jprovazn | do we? | 08:44 |
| lifeless | we're running SSL on the undercloud | 08:44 |
| lifeless | and it's a high pri ticket for the overcloud | 08:44 |
| lifeless | so yeah | 08:44 |
| lifeless | as you know, haproxy can do SSL two ways - tcp proxy or by having a ssl unwrapper in front of it | 08:44 |
| lifeless | But if we unwrap it | 08:45 |
| lifeless | we can't forward over the network. | 08:45 |
| lifeless | If we don't unwrap it, we need a port per service. | 08:45 |
| jprovazn | which we have (port per service) | 08:45 |
| lifeless | but we need two ports per service | 08:45 |
| lifeless | we need one port for haproxy | 08:45 |
| lifeless | and one port for the service | 08:45 |
| lifeless | which you've achieved by manual binding | 08:45 |
| lifeless | at the likely cost of IPv6 support | 08:46 |
| lifeless | so anyhow, my suggestion is: | 08:46 |
| lifeless | - let everything bind as normal, the one exception being keystone's discovery port which we need to override | 08:46 |
| *** ifarkas has quit IRC | 08:46 | |
| lifeless | - pick some arbitrary new ports for each service, have haproxy listen on them | 08:47 |
| lifeless | - and we need to use tcp forwarding, not http, because all of the services will be running SSL before long | 08:47 |
| lifeless | what do you think? | 08:47 |
| lifeless | oh, tweaking that - with keystone, don't bind to a specific ip, jsut change the local port the native keystone service is running on, so that it still binds to * | 08:48 |
| jprovazn | mmt | 08:49 |
| lifeless | and always have haproxy there - it makes it easier to document, plus we can instrument things in one place | 08:49 |
| jprovazn | so keystone returns endpoints to haproxy, right? | 08:50 |
| lifeless | yeah | 08:50 |
| jprovazn | which one? | 08:50 |
| jprovazn | (if VIP is not used) | 08:50 |
| lifeless | this is with VIP | 08:50 |
| lifeless | I'm not suggesting that we rework the design | 08:50 |
| lifeless | I asked some early questions about that | 08:50 |
| *** ifarkas has joined #tripleo | 08:50 | |
| lifeless | but you can ignore them now | 08:50 |
| jprovazn | ah, ok, so if you mean just port binding, that I think your idea makes sense | 08:51 |
| lifeless | so we install keepalived always, we install haproxy always, the pathologic case has nothing to fail over | 08:51 |
| lifeless | we register the vip (well actually the DNS name for the cloud) + ports that haproxy is listening on in keystone | 08:52 |
| lifeless | we could do interesting things in future like updating keystone when failover occurs rather than having haproxy | 08:52 |
| lifeless | but that can wait | 08:52 |
| jprovazn | well, it would be difficult - keystone would have to be notified that something went wrong | 08:53 |
| jprovazn | (which haproxy does pretty well since it's mature) | 08:53 |
| jprovazn | (I hope) | 08:53 |
| lifeless | That why I say future ;) | 08:54 |
| lifeless | it might be a bad idea | 08:54 |
| lifeless | its late here and I'm sleep deprived :( | 08:54 |
| jprovazn | your proposal sounds good, only one thing | 08:54 |
| jprovazn | I wonder if it's really better to use extra ports pointing to haproxy instead of bind openstack services to just local ip | 08:55 |
| jprovazn | I don't see the benefit TBH, I understand that it's alternative which works fine, but don't see the advantage | 08:56 |
| lifeless | ok | 08:56 |
| lifeless | so here are the things I see | 08:56 |
| lifeless | a) operator confusion | 08:57 |
| lifeless | two different tools both listening on port X is really confusing unless you explicitly know to look for it | 08:57 |
| lifeless | b) configuration complexity | 08:57 |
| lifeless | we'll need to manually configure all the local ips - IPv4 and v6, and we haven't done any v6 work so far | 08:58 |
| lifeless | thats really about it | 08:58 |
| lifeless | a) is a big one for me | 08:58 |
| jprovazn | ok, makes sense | 08:58 |
| lifeless | b) is meh - it's just some work to do, but a) would be stuck as part of the design | 08:58 |
| *** rdopieralski has joined #tripleo | 08:59 | |
| lifeless | ok cool. | 08:59 |
| jprovazn | will update the patch, thanks for feedback | 08:59 |
| lifeless | I'm going to go halt() for the night, thanks for your time! | 08:59 |
| jprovazn | good night | 08:59 |
| *** akuznetsov has joined #tripleo | 09:00 | |
| *** jistr has joined #tripleo | 09:03 | |
| *** derekh has joined #tripleo | 09:05 | |
| SpamapS | wait why are we doing haproxy? | 09:09 |
| SpamapS | isn't that lbaas's thing? | 09:09 |
| *** cd-undercloud has joined #tripleo | 09:15 | |
| cd-undercloud | ************** overcloud complete status=42 ************ | 09:15 |
| *** cd-undercloud has quit IRC | 09:15 | |
| Ng | morning all | 09:20 |
| rpodolyaka | o/ | 09:20 |
| *** rushiagr has quit IRC | 09:26 | |
| jprovazn | morning | 09:26 |
| jprovazn | SpamapS: well, haproxy is used for the load balancing | 09:27 |
| *** lucasagomes has joined #tripleo | 09:28 | |
| SpamapS | jprovazn: LoadBalancer as a Service should be doing that. | 09:28 |
| SpamapS | jprovazn: we shouldn't need to invent our own :) | 09:29 |
| SpamapS | https://wiki.openstack.org/wiki/Neutron/LBaaS/HowToRun | 09:29 |
| SpamapS | It seems that it was implemented similar to the L3 agent, so it may not work with our hard wired provider network. | 09:31 |
| jprovazn | SpamapS: thanks, going to look | 09:31 |
| jprovazn | SpamapS: short answer why haproxy is used - as I understand it it's outcome of design summit: https://etherpad.openstack.org/p/tripleo-icehouse-ha-production-configuration also it's the default way suggested in openstack HA guide | 09:32 |
| *** athomas has joined #tripleo | 09:32 | |
| SpamapS | jprovazn: I appreciate that. I think we may end up having to do it if Neutron LBaaS won't fit into our network setup. But we should start with using the aaS's. :) | 09:33 |
| SpamapS | I'd like to eventually be using Trove too. | 09:33 |
| *** marun has quit IRC | 09:36 | |
| SpamapS | jprovazn: https://git.openstack.org/cgit/openstack/neutron/tree/neutron/services/loadbalancer/drivers/haproxy :) | 09:36 |
| *** marun has joined #tripleo | 09:37 | |
| *** bauzas has joined #tripleo | 09:38 | |
| SpamapS | interesting, looks like they just spin haproxy up in a netns right on the network node | 09:40 |
| SpamapS | that doesn't seem like something we could use unfortunately. | 09:43 |
| openstackgerrit | Derek Higgins proposed a change to openstack/tripleo-image-elements: Add extra swift features https://review.openstack.org/61509 | 09:44 |
| jprovazn | SpamapS: didn't get to reasons why we can't use it yet, but I must admit that it looks extremely interesting | 09:46 |
| jprovazn | (at least according to doc :) | 09:46 |
| SpamapS | jprovazn: since we just have a hard wired network for baremetal nodes, it would be challenging to get the traffic into a namespace on the network node.. | 09:48 |
| SpamapS | jprovazn: however, this is highly modular, so I do wonder if we could write one that just works without the netns stuff | 09:49 |
| SpamapS | of course then we have to find a way to get neutron to push that host as the gateway | 09:50 |
| SpamapS | jprovazn: one reason I like doing it with neutron's lbaas though is that vendors are writing hardware backends already | 09:50 |
| SpamapS | anyway.. -> sleep | 09:51 |
| jprovazn | SpamapS: would it make sense to do this in next step (set up haproxy is not so difficult)? | 09:51 |
| jprovazn | good night? | 09:51 |
| jprovazn | s/?/./ | 09:51 |
| SpamapS | jprovazn: yes it is fine to defer using neutron lbaas | 09:53 |
| *** max_lobur_afk is now known as max_lobur | 10:00 | |
| *** rushiagr has joined #tripleo | 10:01 | |
| *** jtomasek has joined #tripleo | 10:05 | |
| *** martyntaylor has joined #tripleo | 10:09 | |
| *** cd-undercloud has joined #tripleo | 10:09 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 10:09 |
| *** cd-undercloud has quit IRC | 10:09 | |
| *** nosnos_ has joined #tripleo | 10:15 | |
| *** nosnos_ has quit IRC | 10:16 | |
| *** nosnos_ has joined #tripleo | 10:16 | |
| *** nosnos has quit IRC | 10:17 | |
| *** nosnos_ has quit IRC | 10:20 | |
| *** rongze has joined #tripleo | 10:41 | |
| *** rongze has quit IRC | 10:46 | |
| *** akuznetsov has quit IRC | 10:55 | |
| *** ProfFalken has joined #tripleo | 10:57 | |
| *** akuznetsov has joined #tripleo | 10:58 | |
| *** ProfFalken has left #tripleo | 10:58 | |
| *** cd-undercloud has joined #tripleo | 11:03 | |
| cd-undercloud | ************** overcloud complete status=1 ************ | 11:03 |
| *** cd-undercloud has quit IRC | 11:03 | |
| *** markmc has joined #tripleo | 11:17 | |
| *** boris-42 has joined #tripleo | 11:17 | |
| *** akuznetsov has quit IRC | 11:19 | |
| *** akuznetsov has joined #tripleo | 11:22 | |
| *** slagle has joined #tripleo | 11:49 | |
| *** cd-undercloud has joined #tripleo | 11:58 | |
| cd-undercloud | ************** overcloud complete status=42 ************ | 11:58 |
| *** cd-undercloud has quit IRC | 11:58 | |
| *** kushi_ has joined #tripleo | 12:05 | |
| kushi_ | I encounter an error while setup devtest env: https://bugs.launchpad.net/tripleo/+bug/1260292 | 12:13 |
| uvirtbot | Launchpad bug 1260292 in tripleo "permission denied error in setup-network" [Undecided,New] | 12:13 |
| kushi_ | Is there some suggestion for this issue ? | 12:14 |
| slagle | kushi_: you shouldn't need sudo there as you're already authenticated with virsh | 12:16 |
| slagle | kushi_: check permissions on /var/lib/libvirt/network/brbm.xml, and the directories leading up to it | 12:17 |
| kushi_ | but the dir is not writable for common user by default, should we add write permission for it ? | 12:20 |
| kushi_ | -rw------- 1 root root 499 Dec 12 03:24 /var/lib/libvirt/network/brbm.xml | 12:21 |
| *** arata has left #tripleo | 12:21 | |
| slagle | that looks fine | 12:22 |
| slagle | drwx------. 2 root root 4096 Dec 4 07:53 /var/lib/libvirt/network/ | 12:22 |
| slagle | odd that you could define the network but not start it | 12:22 |
| kushi_ | run virsh with "sudo" or not will get different result on my Ubuntu | 12:23 |
| kushi_ | drwxr-xr-x 2 root root 4096 Dec 12 03:24 /var/lib/libvirt/network// | 12:23 |
| slagle | ubuntu 12 is pretty old isn't it? would need to get someone more familiar with ubuntu to confirm if it's supposed to work | 12:24 |
| slagle | i think most of the work has been done on raring and after | 12:24 |
| Ng | correct | 12:27 |
| Ng | kushi_: 12.04 definitely won't work right now. You'd need to backport some stuff, but it would be much easier to go with 13.04 or 13.10 | 12:28 |
| kushi_ | Ng: Ubuntu 13.04 is checked ? We'd better add some comment in the wiki page: http://docs.openstack.org/developer/tripleo-incubator/devtest.html | 12:31 |
| Ng | kushi_: we probably should mention that 12.04 isn't supported | 12:34 |
| kushi_ | Ng: agree | 12:35 |
| *** rongze has joined #tripleo | 12:37 | |
| *** bcrochet has quit IRC | 12:38 | |
| * slagle updates bug | 12:40 | |
| *** rongze has quit IRC | 12:41 | |
| *** bcrochet has joined #tripleo | 12:43 | |
| *** cd-undercloud has joined #tripleo | 12:55 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 12:55 |
| *** cd-undercloud has quit IRC | 12:55 | |
| *** sdake_ has joined #tripleo | 12:58 | |
| openstackgerrit | Derek Higgins proposed a change to openstack/tripleo-image-elements: Add tempest element https://review.openstack.org/50477 | 13:01 |
| *** panda has joined #tripleo | 13:01 | |
| *** panda_ has quit IRC | 13:04 | |
| *** jcoufal_ has joined #tripleo | 13:08 | |
| *** jcoufal has quit IRC | 13:10 | |
| *** rongze has joined #tripleo | 13:12 | |
| *** kushi_ has quit IRC | 13:12 | |
| *** rbrady has joined #tripleo | 13:19 | |
| *** rbrady has left #tripleo | 13:20 | |
| *** arata has joined #tripleo | 13:24 | |
| *** arata has left #tripleo | 13:25 | |
| *** rcarrillocruz1 has joined #tripleo | 13:32 | |
| *** rcarrillocruz has quit IRC | 13:33 | |
| *** jdob has joined #tripleo | 13:40 | |
| slagle | jistr: thanks for adding my reply back to the openstack-dev list :) | 13:44 |
| slagle | gmail fail on my part | 13:44 |
| *** cd-undercloud has joined #tripleo | 13:50 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 13:50 |
| *** cd-undercloud has quit IRC | 13:50 | |
| jistr | slagle: hmm i think i didn't add it to openstack-dev, haha :) | 13:50 |
| jistr | slagle: but feel free to forward the whole thing there :) | 13:50 |
| *** jprovazn has quit IRC | 13:51 | |
| *** dprince has joined #tripleo | 14:01 | |
| *** rongze_ has joined #tripleo | 14:01 | |
| *** blamar has quit IRC | 14:01 | |
| *** tzumainn has joined #tripleo | 14:02 | |
| *** rongze has quit IRC | 14:04 | |
| *** jayg|g0n3 is now known as jayg | 14:05 | |
| *** julim has joined #tripleo | 14:09 | |
| openstackgerrit | A change was merged to openstack/tripleo-image-elements: Add extra swift features https://review.openstack.org/61509 | 14:14 |
| *** jprovazn has joined #tripleo | 14:16 | |
| *** athomas has quit IRC | 14:20 | |
| *** rongze has joined #tripleo | 14:20 | |
| *** rongze_ has quit IRC | 14:24 | |
| openstackgerrit | Ralf Haferkamp proposed a change to openstack/diskimage-builder: Merge force-text-mode-console into the bootloader script https://review.openstack.org/61755 | 14:25 |
| *** athomas has joined #tripleo | 14:27 | |
| *** sdake_ has quit IRC | 14:30 | |
| *** jdob has quit IRC | 14:34 | |
| *** jdob has joined #tripleo | 14:34 | |
| *** arata has joined #tripleo | 14:36 | |
| *** cd-undercloud has joined #tripleo | 14:42 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 14:42 |
| *** cd-undercloud has quit IRC | 14:42 | |
| *** rongze_ has joined #tripleo | 14:43 | |
| *** CaptTofu has quit IRC | 14:43 | |
| *** CaptTofu has joined #tripleo | 14:44 | |
| *** rongze has quit IRC | 14:44 | |
| *** blamar has joined #tripleo | 14:45 | |
| *** CaptTofu has quit IRC | 14:49 | |
| dprince | derekh: I'm still seeing intermittent resizefs failures today (after the diskimage-builder fix yesterday) | 14:49 |
| dprince | derekh: similar to the ticket lifeless filed here https://lists.launchpad.net/kernel-packages/msg18566.html | 14:50 |
| *** jergerber has joined #tripleo | 14:51 | |
| derekh | dprince: looks like the error I was seeing , but the fix sorted out the problem for me, hmmm, looks like the size of the target disk is a factor | 14:56 |
| *** martyntaylor1 has joined #tripleo | 14:57 | |
| *** jistr has quit IRC | 14:58 | |
| *** martyntaylor has quit IRC | 14:59 | |
| *** arata has left #tripleo | 14:59 | |
| *** akrivoka has quit IRC | 15:00 | |
| *** akrivoka has joined #tripleo | 15:01 | |
| *** jistr has joined #tripleo | 15:02 | |
| *** jcooley_ has joined #tripleo | 15:05 | |
| *** jtomasek has quit IRC | 15:05 | |
| *** matty_dubs|gone is now known as matty_dubs | 15:08 | |
| *** lsmola has quit IRC | 15:08 | |
| *** lsmola has joined #tripleo | 15:09 | |
| rcarrillocruz1 | hello guys, i'm willing to get my hands on tripleo code, install it and get involved. Where is the code hosted? github? there are quite a few projects with 'tripleO', not sure which ones I should pull... | 15:17 |
| *** sdake_ has joined #tripleo | 15:18 | |
| *** sdake_ has quit IRC | 15:18 | |
| *** sdake_ has joined #tripleo | 15:18 | |
| Ng | rcarrillocruz1: https://wiki.openstack.org/wiki/TripleO has a list :) | 15:18 |
| Ng | rcarrillocruz1: if you want to just get started with a demo TripleO setup, just grab the incubator one and see the docs link next to it, that will end up fetching most of the repos into ~/.cache/tripleo/ anyway | 15:19 |
| Ng | http://docs.openstack.org/developer/tripleo-incubator/devtest.html | 15:19 |
| *** jtomasek has joined #tripleo | 15:19 | |
| rcarrillocruz1 | ah, the incubator fetches other repos as deps | 15:20 |
| rcarrillocruz1 | alright | 15:20 |
| rcarrillocruz1 | thanks! | 15:20 |
| *** rcarrillocruz1 is now known as rcarrillocruz | 15:20 | |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Enable some services after installing daemon's... https://review.openstack.org/59299 | 15:25 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Call os-svc-enable in post-configure.d https://review.openstack.org/59298 | 15:25 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Enable services with os-svc-enable https://review.openstack.org/59297 | 15:25 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Don't enable services in os-svc-install https://review.openstack.org/59300 | 15:25 |
| Ng | rcarrillocruz: well, specifically scripts/devtest.sh does. the incubator repo has a bunch of other stuff in it that's not ready for prime-time yet, but it's most useful for the devtest.sh demo setup :) | 15:26 |
| NobodyCam | good morning TripleO | 15:27 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Enable some services after installing daemon's... https://review.openstack.org/59299 | 15:28 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Call os-svc-enable in post-configure.d https://review.openstack.org/59298 | 15:28 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Enable services with os-svc-enable https://review.openstack.org/59297 | 15:28 |
| openstackgerrit | Dan Prince proposed a change to openstack/tripleo-image-elements: Don't enable services in os-svc-install https://review.openstack.org/59300 | 15:28 |
| Ng | DAN ALL THE THINGS | 15:29 |
| dprince | Ng: morning | 15:35 |
| Ng | hey :) | 15:36 |
| *** cd-undercloud has joined #tripleo | 15:42 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 15:42 |
| *** cd-undercloud has quit IRC | 15:42 | |
| *** ohadlevy has quit IRC | 15:43 | |
| *** vkozhukalov has quit IRC | 15:51 | |
| *** akrivoka has quit IRC | 15:52 | |
| *** akrivoka has joined #tripleo | 15:53 | |
| *** antonym has quit IRC | 15:54 | |
| *** homegrown has quit IRC | 15:55 | |
| *** Guest53164 has joined #tripleo | 16:00 | |
| *** Guest53164 has quit IRC | 16:07 | |
| SpamapS | g'morning | 16:11 |
| openstackgerrit | Ralf Haferkamp proposed a change to openstack/diskimage-builder: Merge force-text-mode-console into the bootloader script https://review.openstack.org/61755 | 16:19 |
| *** jtomasek has quit IRC | 16:24 | |
| *** spzala has joined #tripleo | 16:30 | |
| openstackgerrit | A change was merged to openstack/diskimage-builder: Quieten disk-image-get-kernel https://review.openstack.org/60804 | 16:31 |
| *** bauzas has quit IRC | 16:32 | |
| *** AaronGr_Zzz is now known as AaronGr | 16:33 | |
| *** cd-undercloud has joined #tripleo | 16:38 | |
| cd-undercloud | ************** overcloud complete status=1 ************ | 16:38 |
| *** cd-undercloud has quit IRC | 16:38 | |
| *** jtomasek has joined #tripleo | 16:42 | |
| SpamapS | | fault | {u'message': u'No valid host was found. ', u'code': 500, u'created': u'2013-12-12T12:09:40Z'} | | 16:42 |
| SpamapS | I just wonder if we should spin a few times on that nova boot | 16:42 |
| *** Guest53164 has joined #tripleo | 16:43 | |
| *** martyntaylor1 has quit IRC | 16:44 | |
| *** martyntaylor has joined #tripleo | 16:44 | |
| *** jtomasek has quit IRC | 16:47 | |
| *** boris-42 has quit IRC | 16:48 | |
| *** martyntaylor has quit IRC | 16:49 | |
| *** jistr has quit IRC | 16:50 | |
| *** akrivoka has quit IRC | 16:51 | |
| *** akrivoka has joined #tripleo | 16:52 | |
| *** jtomasek has joined #tripleo | 17:00 | |
| *** rongze_ has quit IRC | 17:01 | |
| *** martyntaylor has joined #tripleo | 17:05 | |
| *** john-n-seattle1 has joined #tripleo | 17:05 | |
| derekh | SpamapS: probably not related to that fault but test-env-testenv0-gvyeaxf2mb7k has been 30 minutes spawning, bad host on undercloud ? PXE deploy doesn't seem to have started | 17:07 |
| *** matty_dubs is now known as matty_dubs|lunch | 17:08 | |
| *** jcoufal_ has quit IRC | 17:10 | |
| *** john-n-seattle1 has quit IRC | 17:12 | |
| rpodolyaka | http://paste.openstack.org/show/54907/ successful rebuild with preserve_ephemeral=True :) | 17:12 |
| *** akrivoka has quit IRC | 17:13 | |
| rpodolyaka | lifeless: ^ update on rebuild | 17:13 |
| *** jtomasek has quit IRC | 17:15 | |
| SpamapS | derekh: baremetal will give up on one host and move on to another | 17:17 |
| *** lsmola has quit IRC | 17:18 | |
| derekh | ok | 17:18 |
| derekh | SpamapS: thanks | 17:18 |
| SpamapS | derekh: nova-compute.log on cd-undercloud might have clues | 17:19 |
| *** max_lobur is now known as max_lobur_afk | 17:19 | |
| derekh | SpamapS: ahh yes, it seems to have timeout and started again | 17:20 |
| SpamapS | derekh: those boxes take 7-8 minutes per POST, so 16 minutes of POST alone is normal. | 17:21 |
| derekh | ok | 17:21 |
| SpamapS | I do wish we could somehow link up console-log with the IPMI console | 17:23 |
| *** markmc has quit IRC | 17:24 | |
| openstackgerrit | Derek Higgins proposed a change to openstack-infra/tripleo-ci: Update toci_gate_test.sh to use test_env https://review.openstack.org/61829 | 17:26 |
| openstackgerrit | Derek Higgins proposed a change to openstack-infra/tripleo-ci: Adding testenv worker wrapper script https://review.openstack.org/61052 | 17:26 |
| *** john-n-seattle1 has joined #tripleo | 17:28 | |
| derekh | dprince: pleia2 ^^ so I had a test env working earlier this those working, dprince I also have an update for the heat template you proposed I'll send it to you so you can use it to help update your changeset | 17:30 |
| *** cd-undercloud has joined #tripleo | 17:31 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 17:31 |
| *** cd-undercloud has quit IRC | 17:31 | |
| *** martyntaylor has quit IRC | 17:31 | |
| *** rdopieralski has quit IRC | 17:32 | |
| *** john-n-seattle1 has left #tripleo | 17:32 | |
| *** blamar has quit IRC | 17:33 | |
| pleia2 | derekh: great | 17:34 |
| pleia2 | derekh: have some time for you and dprince to sync up with me on where we stand? | 17:34 |
| derekh | pleia2: I have about 20 minutes now if you want and dprince is around | 17:35 |
| pleia2 | derekh: now is good :) | 17:35 |
| derekh | dprince: your there ? | 17:35 |
| dprince | derekh: yes | 17:36 |
| derekh | wanna chat, quick status update | 17:36 |
| dprince | pleia2/derekh: sure. I'll start a hangout... | 17:36 |
| derekh | cool | 17:36 |
| Ng | dprince: might be worth talking to lifeless about his -2 on 59298 | 17:37 |
| dprince | Ng: will do | 17:39 |
| Ng | dprince: it seems like it can probably come off, but atm it's holding up all your changesets above it | 17:40 |
| *** gothicmindfood has joined #tripleo | 17:43 | |
| *** gothicmindfood has left #tripleo | 17:43 | |
| SpamapS | dprince: FWIW, now that I've been staring at all of this code this week, I understand why your noop would have been fine. | 17:54 |
| SpamapS | dprince: apologies for blocking your progress while I caught up. | 17:54 |
| derekh | dprince: pleia2 this is the command I was using to build a test-env image | 17:55 |
| derekh | ELEMENTS_PATH=/root/tmp/tripleo-image-elements/elements:tripleo-ci/elements /opt/stack/diskimage-builder/bin/disk-image-create -o testenv local-config stackuser ubuntu serial-console dhcp-all-interfaces testenv-worker | 17:55 |
| *** ohadlevy has joined #tripleo | 17:56 | |
| *** bauzas has joined #tripleo | 17:58 | |
| *** athomas has quit IRC | 17:58 | |
| *** derekh has quit IRC | 17:58 | |
| SpamapS | Cores: https://review.openstack.org/#/c/59297/ wants your +2 | 17:59 |
| SpamapS | https://review.openstack.org/#/c/60341/9 too | 17:59 |
| dprince | SpamapS: no worries, I don't *really* care about the order. Just trying to do what is easiest | 18:05 |
| SpamapS | lifeless: https://review.openstack.org/#/c/59298/ <-- definitely ready for you to reevaluate.. I think the -2 comments are entirely addressed and this is ready to go now. | 18:05 |
| SpamapS | dprince: glad to hear it. Nonetheless, unintended consequence of my misunderstanding is that you have been delayed on this change for quite a while. | 18:06 |
| *** vkozhukalov has joined #tripleo | 18:12 | |
| *** matty_dubs|lunch is now known as matty_dubs | 18:14 | |
| *** rpodolyaka1 has joined #tripleo | 18:17 | |
| *** lucasagomes has quit IRC | 18:24 | |
| *** cd-undercloud has joined #tripleo | 18:26 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 18:26 |
| *** cd-undercloud has quit IRC | 18:26 | |
| *** krotscheck has joined #tripleo | 18:37 | |
| *** rcarrillocruz1 has joined #tripleo | 18:37 | |
| *** rcarrillocruz has quit IRC | 18:37 | |
| lifeless | SpamapS: dprince ack | 18:47 |
| *** rushiagr has quit IRC | 18:47 | |
| dprince | lifeless: morning. | 18:48 |
| dprince | lifeless: hey, I'm still hitting resizefs issues today. Lots of them. Similar to what you filed here: https://lists.launchpad.net/kernel-packages/msg18566.html | 18:48 |
| dprince | lifeless: I'm actually trying to get some more info on it because when it happens my system (bare metal) is totally hosed... | 18:48 |
| dprince | lifeless: currently, I'm building an image where the code in cloud-init to actually run the resizefs is just disabled. Then my plan was to log in and run it manually to see if I can get more info that way. Wondering if you had any more ideas on this one? | 18:49 |
| *** blamar has joined #tripleo | 18:59 | |
| lifeless | dprince: not offhand | 19:06 |
| lifeless | dprince: but - you might try using flavour that has a smallish / and a large ephemeral | 19:06 |
| *** krotscheck has quit IRC | 19:12 | |
| *** ohadlevy has quit IRC | 19:16 | |
| *** ohadlevy has joined #tripleo | 19:17 | |
| *** ohadlevy is now known as Guest5361 | 19:17 | |
| *** cd-undercloud has joined #tripleo | 19:23 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 19:23 |
| *** cd-undercloud has quit IRC | 19:23 | |
| *** Guest5361 has quit IRC | 19:27 | |
| *** ohadlevy_ has joined #tripleo | 19:29 | |
| *** vipul is now known as vipul-away | 19:30 | |
| *** rcarrillocruz1 has quit IRC | 19:41 | |
| *** vipul-away is now known as vipul | 19:44 | |
| *** ohadlevy_ has quit IRC | 19:48 | |
| *** akuznetsov has quit IRC | 19:49 | |
| *** ohadlevy_ has joined #tripleo | 19:49 | |
| *** ohadlevy_ has quit IRC | 19:58 | |
| *** jprovazn has quit IRC | 19:58 | |
| *** ohadlevy_ has joined #tripleo | 20:00 | |
| *** akuznetsov has joined #tripleo | 20:09 | |
| *** vipul is now known as vipul-away | 20:13 | |
| *** vipul-away is now known as vipul | 20:13 | |
| *** jcoufal has joined #tripleo | 20:14 | |
| lifeless | jcoufal: morning! well, evening! | 20:15 |
| lifeless | tzumainn: and hi | 20:15 |
| jcoufal | lifeless: hey | 20:15 |
| jcoufal | lifeless: tzumainn wrote that he might skip actually | 20:16 |
| lifeless | ok | 20:16 |
| jcoufal | tzumainn: however if you want to join you are welcome | 20:16 |
| *** ohadlevy_ has quit IRC | 20:19 | |
| *** cd-undercloud has joined #tripleo | 20:20 | |
| cd-undercloud | ************** overcloud complete status=42 ************ | 20:20 |
| *** cd-undercloud has quit IRC | 20:20 | |
| *** ohadlevy has joined #tripleo | 20:22 | |
| *** ohadlevy is now known as Guest52964 | 20:22 | |
| *** vipul is now known as vipul-away | 20:28 | |
| tzumainn | jcoufal, I don't think I need to, based on what we talked about earlier | 20:33 |
| *** UtahDave has joined #tripleo | 20:33 | |
| *** bauzas has quit IRC | 20:44 | |
| *** vipul-away is now known as vipul | 20:49 | |
| *** jcooley_ has quit IRC | 20:49 | |
| *** cwolferh has joined #tripleo | 20:52 | |
| lifeless | jcoufal: ironic has a key:value field per node | 20:53 |
| lifeless | jcoufal: so we can say rack=xyz in that field if we want. | 20:53 |
| jcoufal | lifeless: which is actually what we need | 20:53 |
| jcoufal | lifeless: that's great news | 20:53 |
| lifeless | jcoufal: it doesn't support searching it yet though, but yeah - we can stash per node data there as needed | 20:53 |
| lifeless | jcoufal: so the 'tags' field can let the user put what they want as 'foo' or as 'foo=bar', and we can store 'foo' as 'foo=""' | 20:54 |
| lifeless | IMO | 20:54 |
| jcoufal | awesome | 20:54 |
| jcoufal | we can handle the search at the UI part - at least some basic filetring | 20:55 |
| jcoufal | lifeless: thanks for checking | 20:55 |
| lifeless | anytime :) | 20:56 |
| *** bauzas has joined #tripleo | 21:01 | |
| *** rpodolyaka1 has quit IRC | 21:03 | |
| *** boris-42 has joined #tripleo | 21:03 | |
| tzumainn | jcoufal, does that let us find nodes with particular roles more easily? | 21:04 |
| *** ccrouch has quit IRC | 21:05 | |
| jcoufal | tzumainn: we can tag anything. so for the first iteration it might be useful for hardcoded roles | 21:05 |
| *** rwsu has quit IRC | 21:05 | |
| *** bauzas has quit IRC | 21:06 | |
| tzumainn | jcoufal, awesome - the only issue I'd potentially have is that - what happens if a node is manipulated outside of tuskar? then the tag can become inaccurate | 21:06 |
| jcoufal | tzumainn: like somebody ssh to the node and change things manually? | 21:07 |
| tzumainn | or someone shuts down the service on the node manually, or something | 21:07 |
| jcoufal | tzumainn: well we will be checking the services if they are running anyway - rising warnings | 21:08 |
| jcoufal | tzumainn: however for the first steps I wouldn't complicate things that much | 21:09 |
| *** epim has joined #tripleo | 21:09 | |
| tzumainn | jcoufal, right, it just seems that we need to be careful, in that a tag on a node in ironic isn't a guarantee that the tag is correct | 21:09 |
| tzumainn | jcoufal, are you going to send out a summary of your conversation with lifeless? | 21:09 |
| lifeless | yes | 21:10 |
| lifeless | mmm, don't conflate tags - user data - with stuff we need | 21:10 |
| lifeless | we can ask heat for what instances we have at any point in time | 21:10 |
| lifeless | no need for tags | 21:10 |
| jcoufal | tzumainn: yes, mostly trying to wrap up all the conversations, just not tonight, tomorrow morning | 21:10 |
| tzumainn | lifeless, yeah, that makes sense to me | 21:11 |
| SpamapS | tzumainn: That is why arbitrary hard coded roles are problematic | 21:12 |
| SpamapS | tzumainn: (re "what if something manipulates the node outside ironic") | 21:13 |
| tzumainn | SpamapS, yeah, agreed - I think what I was trying to get to in a roundabout way was that I don't think we should rely on ironic tags for stuff like that | 21:14 |
| SpamapS | :) | 21:14 |
| *** cd-undercloud has joined #tripleo | 21:16 | |
| cd-undercloud | ************** overcloud complete status=1 ************ | 21:16 |
| *** cd-undercloud has quit IRC | 21:16 | |
| tzumainn | lifeless, jcoufal, out of curiosity, was there agreement to push manual node allocation to a future consideration? | 21:17 |
| jcoufal | tzumainn: it will be ongoing discussion | 21:18 |
| jcoufal | so basically yes | 21:18 |
| tzumainn | jcoufal, okay, I think that as far as the requirements discussion, that was the biggest sticking point, so that's good to hear | 21:18 |
| jcoufal | tzumainn: need to run, it's getting late | 21:19 |
| *** ccrouch has joined #tripleo | 21:19 | |
| tzumainn | jcoufal, definitely, good night! | 21:19 |
| jcoufal | c'you tomorrow guys | 21:19 |
| *** krotscheck has joined #tripleo | 21:36 | |
| openstackgerrit | Clint "SpamapS" Byrum proposed a change to openstack/tripleo-incubator: Wait for Neutron L2 Agent on Compute Node https://review.openstack.org/61608 | 21:36 |
| *** spzala has quit IRC | 21:37 | |
| openstackgerrit | Ben Nemec proposed a change to openstack/tripleo-image-elements: Add tempest element https://review.openstack.org/50477 | 21:44 |
| *** marun has quit IRC | 21:48 | |
| *** jdob has quit IRC | 21:50 | |
| *** cwolferh has quit IRC | 21:54 | |
| *** julim_ has joined #tripleo | 22:05 | |
| *** julim has quit IRC | 22:07 | |
| *** cd-undercloud has joined #tripleo | 22:11 | |
| cd-undercloud | ************** overcloud complete status=42 ************ | 22:11 |
| *** cd-undercloud has quit IRC | 22:11 | |
| *** boris-42 has quit IRC | 22:11 | |
| *** vkozhukalov has quit IRC | 22:15 | |
| *** dprince has quit IRC | 22:22 | |
| *** matty_dubs is now known as matty_dubs|gone | 22:32 | |
| *** noslzzp has quit IRC | 22:33 | |
| *** phschwartz is now known as phschwartz_aw | 22:46 | |
| *** phschwartz_aw is now known as phschwartz | 22:47 | |
| *** jayg is now known as jayg|g0n3 | 22:59 | |
| *** jergerber has quit IRC | 23:02 | |
| *** cd-undercloud has joined #tripleo | 23:07 | |
| cd-undercloud | ************** overcloud complete status=0 ************ | 23:07 |
| *** cd-undercloud has quit IRC | 23:07 | |
| *** phschwartz has quit IRC | 23:10 | |
| *** phschwartz has joined #tripleo | 23:11 | |
| *** jcoufal has quit IRC | 23:12 | |
| openstackgerrit | A change was merged to openstack/tripleo-incubator: Wait for Neutron L2 Agent on Compute Node https://review.openstack.org/61608 | 23:36 |
| *** krotscheck has quit IRC | 23:49 | |
| *** ccrouch has quit IRC | 23:51 | |
| lifeless | pleia2: nuts, was today our sync up day? or yesterday? | 23:52 |
| openstackgerrit | Clint "SpamapS" Byrum proposed a change to openstack/tripleo-incubator: Wait for nova-compute to claim to be up https://review.openstack.org/61894 | 23:57 |
| SpamapS | lifeless: ^^ ok that is the last spurious error thing I will work on.. unless more come up. :) | 23:57 |
| SpamapS | lifeless: now, on to preserve_ephemeral! | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!