| *** pbourke_ has quit IRC | 00:13 | |
| *** pbourke_ has joined #heat | 00:13 | |
| *** hongbin has quit IRC | 01:45 | |
| *** hongbin has joined #heat | 03:05 | |
| *** hongbin has quit IRC | 04:00 | |
| *** ramishra has joined #heat | 04:34 | |
| *** zzzeek has quit IRC | 04:41 | |
| *** zzzeek has joined #heat | 04:41 | |
| *** ricolin has joined #heat | 04:45 | |
| *** Chaserjim has joined #heat | 05:02 | |
| *** gfidente has joined #heat | 07:20 | |
| *** rcernin has quit IRC | 07:23 | |
| *** gkadam has joined #heat | 07:52 | |
| *** shardy has joined #heat | 08:14 | |
| *** jtomasek has joined #heat | 08:57 | |
| lxkong | hi heat, if i defined multiple OS::Heat::SoftwareDeployment(no dependencies between them) for a single vm, is it true that those deployment are executed in sequence? That's what i saw from heat-container-agent log. | 09:10 |
|---|---|---|
| *** k_mouza has joined #heat | 09:15 | |
| therve | lxkong: I believe so | 09:24 |
| therve | Agents don't do parallelism | 09:24 |
| lxkong | therve: thanks for the answer. Is there any heat approach for parallel actions inside the vm? | 09:25 |
| therve | lxkong: You can reimplement agents | 09:25 |
| therve | But no not currently | 09:25 |
| lxkong | therve: ok | 09:25 |
| * lxkong is sad for SoftwareDeployment design | 09:26 | |
| lxkong | therve: thanks anyway | 09:26 |
| therve | lxkong: You shouldn't use SoftwareDeployment | 09:27 |
| therve | it's bad | 09:27 |
| lxkong | it's not *me* who is using SoftwareDeployment, it's Magnum which I'm working on | 09:28 |
| therve | Make magnum uses something else | 09:28 |
| therve | lxkong: Instead of having multiple deployment, just have one | 09:33 |
| therve | And wrap your action in there to run them in parallel | 09:33 |
| strigazi | therve: when you say "You shouldn't use SoftwareDeployment", "it's bad", you mean we should never use them? | 09:36 |
| therve | strigazi: Yeah | 09:36 |
| strigazi | therve: is there an alternative option in heat? | 09:37 |
| therve | strigazi: No | 09:37 |
| therve | It's just my opinion, not an official statement | 09:39 |
| strigazi | therve: Got it. What are the arguments against it? | 09:40 |
| therve | strigazi: I think the pull mechanims is inherently incompatible with Heat | 09:43 |
| therve | Nodes just don't communicate back and there is nothing you can do about it | 09:43 |
| therve | Except wait and timeout | 09:43 |
| therve | And then all your deployment breaks because one node lost connectivity | 09:44 |
| therve | It's also made worse by Heat not supporting partial success | 09:44 |
| *** k_mouza has quit IRC | 09:45 | |
| *** k_mouza has joined #heat | 09:46 | |
| strigazi | I have inentified the lack of partial success my self as well as major pain point. Haven't thought that the polling design is orthogonal to heat. | 09:47 |
| ricolin | strigazi, yep, that's what you shared in user feedback in Heat before, and why we triggered discussion in Rocky PTG https://etherpad.openstack.org/p/ptg-rocky-heat-partial-success | 09:52 |
| ricolin | strigazi, No one is working on that for now | 09:53 |
| *** brtknr has joined #heat | 11:03 | |
| brtknr | hey all | 11:04 |
| brtknr | hi all, what do I do if my heat cluster refuses to be deleted? i can only think of deleting it from the database ... | 11:04 |
| *** pbourke_ has quit IRC | 11:05 | |
| therve | brtknr: Identify why it refuses. Otherwise, the database. | 11:06 |
| brtknr | http://paste.openstack.org/show/736888/ | 11:07 |
| brtknr | here's the output of openstack stack event list | 11:08 |
| *** pbourke_ has joined #heat | 11:08 | |
| brtknr | the deletion gets cancelled internally as soon as it is invoked | 11:08 |
| *** k_mouza has quit IRC | 11:09 | |
| *** sshnaidm has quit IRC | 11:15 | |
| *** sshnaidm has joined #heat | 11:16 | |
| *** openstackgerrit has joined #heat | 11:29 | |
| openstackgerrit | Sorin Sbarnea proposed openstack/heat-templates master: Make heat-config logging more human friendly https://review.openstack.org/623516 | 11:29 |
| *** k_mouza has joined #heat | 11:30 | |
| *** jtomasek has quit IRC | 11:34 | |
| *** jtomasek has joined #heat | 12:17 | |
| *** jistr is now known as jistr|medchk | 12:57 | |
| *** k_mouza has quit IRC | 13:02 | |
| *** jtomasek_ has joined #heat | 13:14 | |
| *** jtomasek has quit IRC | 13:15 | |
| *** jtomasek_ is now known as jtomasek | 13:19 | |
| *** k_mouza has joined #heat | 13:20 | |
| *** k_mouza has quit IRC | 13:26 | |
| *** k_mouza has joined #heat | 13:27 | |
| *** k_mouza_ has joined #heat | 13:29 | |
| *** k_mouza has quit IRC | 13:32 | |
| *** jmlowe has quit IRC | 13:45 | |
| *** jistr|medchk is now known as jistr | 13:56 | |
| *** hjensas has quit IRC | 14:13 | |
| *** openstackstatus has joined #heat | 14:17 | |
| *** ChanServ sets mode: +v openstackstatus | 14:17 | |
| *** k_mouza has joined #heat | 14:22 | |
| *** jmlowe has joined #heat | 14:24 | |
| *** jmlowe has quit IRC | 14:24 | |
| *** gkadam has quit IRC | 14:24 | |
| *** k_mouza_ has quit IRC | 14:25 | |
| *** jmlowe has joined #heat | 14:25 | |
| *** jmlowe has quit IRC | 14:30 | |
| *** jmlowe has joined #heat | 14:35 | |
| *** gfidente has quit IRC | 14:49 | |
| *** gfidente has joined #heat | 14:53 | |
| openstackgerrit | Sorin Sbarnea proposed openstack/heat-agents master: Make heat-config logging more human friendly https://review.openstack.org/624112 | 15:44 |
| openstackgerrit | Sorin Sbarnea proposed openstack/heat-templates master: Make heat-config logging more human friendly https://review.openstack.org/623516 | 15:45 |
| *** jmlowe has quit IRC | 15:46 | |
| *** hjensas has joined #heat | 16:08 | |
| *** gmoro has quit IRC | 16:11 | |
| *** gmoro has joined #heat | 16:14 | |
| *** jmlowe has joined #heat | 16:24 | |
| *** jmlowe has quit IRC | 16:34 | |
| *** openstackgerrit has quit IRC | 16:35 | |
| *** k_mouza_ has joined #heat | 16:50 | |
| *** gfidente has quit IRC | 16:51 | |
| *** k_mouza has quit IRC | 16:53 | |
| *** k_mouza_ has quit IRC | 16:55 | |
| *** gfidente has joined #heat | 16:56 | |
| *** hjensas has quit IRC | 17:03 | |
| *** ricolin has quit IRC | 17:05 | |
| *** ricolin has joined #heat | 17:14 | |
| *** jmlowe has joined #heat | 18:12 | |
| *** ricolin has quit IRC | 18:22 | |
| *** openstackgerrit has joined #heat | 18:30 | |
| openstackgerrit | Merged openstack/heat master: Delete db resources not in template https://review.openstack.org/609962 | 18:30 |
| *** jcoufal has joined #heat | 19:01 | |
| *** gfidente is now known as gfidente|adk | 19:10 | |
| *** gfidente|adk is now known as gfidente|afk | 19:10 | |
| *** hjensas has joined #heat | 19:27 | |
| *** Blotis has joined #heat | 19:27 | |
| Blotis | Test | 19:27 |
| Blotis | I was wondering if someone could let me know what are the main differences between Heat and the Openstack SDK....and more specifically, how does one know which one to chose? | 19:29 |
| therve | Blotis: Heat is a service with an API, and the main interface is through templates | 19:54 |
| therve | The SDK is a library | 19:54 |
| therve | Heat is bullt upon the SDK | 19:55 |
| Blotis | I get it. Now, are there any particular cases in which you want to use the SDK instead of Heat? | 20:00 |
| therve | It depends on what you need really | 20:05 |
| therve | But sure there are cases where the sdk makes more sense, if you want more control/flexibility | 20:06 |
| Blotis | hmm ok | 20:08 |
| Blotis | Lets say I want to create my own specific modeling language to model i.e. very specific scenarios, for easy deployment (consider it an abstraction layer on top of heat, sdk and configuration management tools). Now, i want to make it platform independent, which means that I should be able to deploy my scenario in e.g. Openstack, esxi, aws etc... In this particular case, when it comes to Openstack, it makes more sense to rely on the SKD | 20:11 |
| Blotis | than Heat, right? | 20:11 |
| Blotis | (I exclude terraform for now) | 20:13 |
| therve | Are you going to use CFN to talk to AWS? | 20:25 |
| *** jmlowe has quit IRC | 20:26 | |
| *** spotz has joined #heat | 20:29 | |
| *** spotz has left #heat | 20:30 | |
| *** jmlowe has joined #heat | 20:32 | |
| *** jmlowe has quit IRC | 20:33 | |
| *** jmlowe has joined #heat | 20:58 | |
| *** gfidente|afk has quit IRC | 21:03 | |
| *** jtomasek has quit IRC | 21:32 | |
| Blotis | never mind | 21:44 |
| Blotis | thanks anyways | 21:44 |
| *** Blotis has left #heat | 21:44 | |
| zaneb | lxkong: my advice on parallel software deployments would be to write shorter scripts. preferably as short as "docker run ..." | 22:59 |
| *** rcernin has joined #heat | 22:59 | |
| *** shardy has quit IRC | 23:29 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!