16:00:45 #startmeeting Network Meeting 16:00:45 Meeting started Wed Mar 22 16:00:45 2017 UTC. The chair is funzo. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:45 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:00:45 The meeting name has been set to 'network_meeting' 16:01:19 #topic Networking Update 16:01:45 #topic RC2 is planned for today 16:02:48 Show of hands, who's tested RC1? 16:03:26 standing agenda: https://github.com/ansible/community/issues/110 16:04:17 I'll wait a couple more minutes to see if folks trickle in 16:05:08 a couple of the networking core team members are out today, so we might have a lightly attended meeting 16:05:17 Has anyone here tested RC1? 16:06:30 #chair andriusb 16:06:30 Current chairs: andriusb funzo 16:06:37 #chair grybak 16:06:37 Current chairs: andriusb funzo grybak 16:06:53 alrighty then, let's touch on the standing agenda... 16:07:30 take a look at the last comment of https://github.com/ansible/community/issues/110 16:08:28 #info RC2 comes out today (planned), but in case you haven't seen the RC1 details and want to test it now https://groups.google.com/forum/#!topic/ansible-devel/V2ESSQqLnS0 16:08:55 #chair lbalhar 16:08:55 Current chairs: andriusb funzo grybak lbalhar 16:09:20 lbalhar: the meeting started a few minutes ago and we are just taking a look at the standing agenda. Have you had a chance to test RC1? 16:09:23 RC2 is coming out today 16:10:42 Ok if no other comments, look for an email to the dev mailing list for the RC2 announcement today 16:10:50 #topic Open Floor 16:11:22 I'll start the timer for 60 seconds, then move on 16:12:49 60 seconds have passed and I will now go continue to sit alone in the dark 16:12:55 thank you and good day 16:13:23 privateip: you joined at the very end, anything to add other than RC2 planned for today? 16:13:31 very end because i'm alone in the dark 16:13:49 thats what i call timing :) 16:13:58 o/*\o 16:14:02 Q&A session in an hour 16:14:16 oops 16:14:17 two hours 16:14:38 do you have the info handy to paste it here? 16:14:39 and the wolves are after him 16:14:47 https://attendee.gotowebinar.com/register/8226671723620160257 16:14:58 https://www.ansible.com/ask-ansible 16:15:05 #info Ask the Experts session ^^ 16:15:14 thanks for the reminder 16:15:18 any and all questions welcome and fair game 16:15:49 For anyone that has joined late, we now have an open floor 16:16:03 and you'll be able to read the history on the meetbot link posted at the end 16:16:22 ok, any other items to bring up to the group? 16:16:48 * skg-net waves 16:16:58 o/ 16:17:00 hi skg-net 16:17:07 I would like discuss on Dell modules for Ansible 2.3 16:17:16 privateip: hi 16:17:26 skg-net: ok, cool - what specifically would you like to talk about 16:18:38 I have submitted the PR for dellos9 and dellos10 https://github.com/ansible/ansible/pull/22856 16:19:13 would like to get it merged.. 16:19:48 privateip: have you looked over this, I see gundalow did 16:20:01 i haven't but i have some time this afternoon to review it 16:20:55 * privateip commits to reviewing it in depth and comments (or merging) today 16:20:56 As part of the testing we found that, the facts for a CLI previously use to return..in a single item of the list.. 16:21:14 Thanks @privateip 16:21:15 thats a big PR 16:21:16 :) 16:21:22 np 16:22:15 skg-net: thanks for your contribution! 16:22:45 Now, CLI show output is large, its splitting into multiple lines... 16:22:47 skg-net: if you could watch it closely as it gets reviewed, we are really coming down to the wire on releasing 2.3 so any changes requested would need to be made quickly 16:23:00 that will make it go much faster 16:23:01 is this expected 16:23:23 no that is definitely not expected 16:23:49 Eg: For two elements in command list the responses list has three elements .i.e here the second command output is split and stored. commands = [‘show version | display-xml’, ‘show system | display-xml’] responses = ['\n\n \n … \n’, '\n\n \n… ', 'fail\n 16:24:25 yeah that seems odd 16:24:30 sure @funzo, will respond to comments as soon as possible.. 16:24:34 i suspect its a false positive on the prompt match 16:25:14 privateip: The code is mostly what you have written..:) 16:25:35 ok, will debug and keep you posted on that.. 16:25:40 oh in that case its correct... i never introduce bugs :) 16:26:12 *crickets* 16:26:15 lol 16:26:30 that was hard to type with a straight ummm face ;) 16:26:39 i got a great laugh out of it 16:27:01 any other detail on that issue to discuss here? 16:27:14 not until i can dig into it more this PM 16:27:14 How do we need to handle platform specific items in the new connection plugin ? 16:27:31 what do you mean? 16:27:38 do you have an example? 16:29:05 the terminal length 0 can be executed only when the device is authorized for dellos6 16:29:48 but in the current infra, its handled before the authorize.. 16:31:25 hmmmm 16:31:30 dowe write a new connection plugin in such case? 16:31:56 its just an example..dellos6 deviates more.. 16:32:14 there is an on_authorize hook in the terminal plugin 16:32:24 that would allow you to run the commands post authorize 16:33:29 let me explore that option.. 16:33:53 kk 16:34:16 alrighty, moving on then... any other issues to bring up to the group? 16:34:30 Last item..we also hit randomly the command output is partial sometimes.. 16:35:23 this is with dellos9_command module.. 16:35:37 is there an open issue? 16:35:43 https://github.com/ansible/ansible/issues/22863 16:35:44 like that? 16:35:48 except not ios 16:36:22 yes thats correct 16:36:44 we haven't opened issues yet.. 16:37:01 please do for tracking purposes 16:37:10 sure, will do.. 16:37:18 thanks 16:37:43 ok, if there are no other items to bring up to the group I'll pull the plug 16:37:50 one sec 16:37:55 kennc: o/ 16:38:17 with the new net_cli, what happens if the sessions fails between tasks? 16:38:36 you mean if the connection drops? 16:39:12 correct, or times out because the task before takes too long for all the hosts to complete 16:39:27 the next task will create a new connection 16:39:30 and can you confirm there is no going back to the old method if net_cli has issues? 16:39:39 thanks for that info! 16:40:10 did that answer both of your questions? 16:40:12 also, if you know you are going to be running long tasks, there is a timeout you can alter just as an fyi 16:40:18 yes 16:40:21 no, just the first 16:40:51 oh you want to know if there is a way to go back to the old method 16:41:11 yes 16:41:23 sorry im not following... whats the "old method"? 16:41:31 not creating a socket 16:42:39 and one more to add on, how to use delgate_to? Trying to wrap my head around ansible_ssh_user, I feel like that both the jump host and network device wants to use that 16:42:55 no going back from the new arch 16:42:59 no going back from the new architecture 16:43:29 k, I like definitive answers, even when I don't agree :) 16:43:29 you can just say 'ngbftna' and everyone will get it 16:43:54 lol 16:44:08 lol 16:44:12 as far as delegate_to ... it doesn't work with the new connection plug 16:44:26 now, i'll go back to my second question :) 16:44:44 you see the issue here? 16:45:01 but you can use ProxyCommand as documented 16:45:04 http://docs.ansible.com/ansible/faq.html#how-do-i-configure-a-jump-host-to-access-servers-that-i-have-no-direct-access-to 16:45:12 to achieve the same thing but in a more secure way 16:45:20 ok.... that is the piece I was missing 16:45:26 let me look at this 16:46:01 can you add an action to add this to gundalow's documentation ? 16:47:07 and potentially provide an example? not %100 making sense, but it sounds to be what I am looking for 16:47:07 #action Add ProxyCommand reference for how to more securely replace delagate_to with net_cli 16:47:23 thanks 16:47:53 and I still still having issues with ios_banner, but rcarrillocruz is working on it 16:48:01 that's it from me 16:48:37 ok anyone else have anything for the group? 16:49:42 thanks everyone for your participation 16:50:10 #endmeeting