*** jamesmcarthur has joined #openstack-uc | 03:13 | |
*** jamesmcarthur has quit IRC | 04:18 | |
*** jamesmcarthur has joined #openstack-uc | 07:18 | |
*** jamesmcarthur has quit IRC | 07:23 | |
*** VW has joined #openstack-uc | 12:43 | |
*** ChanServ sets mode: +o VW | 12:43 | |
*** jamesmcarthur has joined #openstack-uc | 13:47 | |
*** jamesmcarthur has quit IRC | 13:47 | |
*** jamesmcarthur has joined #openstack-uc | 13:48 | |
*** shintaro has joined #openstack-uc | 14:02 | |
*** shintaro has quit IRC | 15:00 | |
*** VW_ has joined #openstack-uc | 15:25 | |
*** VW_ has quit IRC | 15:27 | |
*** VW_ has joined #openstack-uc | 15:27 | |
*** VW has quit IRC | 15:29 | |
spotz | Nice mail jamesmcarthur | 15:44 |
---|---|---|
jamesmcarthur | :) thanks spotz: | 15:45 |
*** VW_ has quit IRC | 16:05 | |
*** VW has joined #openstack-uc | 16:06 | |
*** ChanServ sets mode: +o VW | 16:06 | |
mrhillsman | i have a concern that i am trying to figure out the right words re that thread | 16:32 |
mrhillsman | re the branding, chris (ops meetup team) mentions often that he has issue with some wording and feeling like it is not for him | 16:34 |
spotz | mrhillsman: That's why I liked Thierry's use of contributer | 16:34 |
mrhillsman | in particular | 16:34 |
mrhillsman | "Who Should Not Attend | 16:34 |
mrhillsman | 09:39 The event is not optimized for non-contributors or people who can’t relate to a specific project team. Each team is given a single room and attendees are expected to spend all their time with their project team. Attendees who can’t pick a specific team to work with are encouraged to skip the event in favor of attending the OpenStack Summit, where a broader range of topics is discussed." | 16:34 |
mrhillsman | agreed and i was going to comment, but wanted to hold off, and ask something here | 16:35 |
mrhillsman | because of that | 16:35 |
mrhillsman | it made me think again back to what VW has been saying | 16:35 |
mrhillsman | what software should do + how software should do it | 16:36 |
spotz | I guess working most closely with OSA we are very heavy on working with OPs and their feedback | 16:36 |
mrhillsman | and i think we have to be careful - correct me if i am wrong or i would like comments - of the "ptg" being discussion about "what software should do" | 16:36 |
mrhillsman | thierry mentioned some concern around the ptg also becoming less focused on actual work as initially intended | 16:37 |
mrhillsman | kind of in passing when we were in dublin | 16:38 |
spotz | I always took PTG (and keep in mind I haven't attended but did attend a couple mid-cycles) this as what should our priorities be for this cycle and how can we meet them | 16:38 |
spotz | And then fleshing things out. This from attending OSA and Barbican only though | 16:39 |
VW | no ptg (or the actual part of teams in rooms around specific projects) should be "how we should make the software work" | 16:39 |
mrhillsman | from being there, at least from my perspective and what i have been exposed to, first 2 days are that, when the other three days being getting started on work where you can | 16:39 |
VW | forum sessions with dev+ops are "what should the software do" | 16:39 |
mrhillsman | sorry if i am getting them mixed up | 16:39 |
mrhillsman | s/if// | 16:40 |
VW | well, those are technically my view of the world put in to words | 16:40 |
VW | so nothing official about them :) | 16:40 |
mrhillsman | i agree, i like trying to boil things down to the basics as well | 16:40 |
VW | but yes, if the events are combined, then we need to honor the intent that a large part of what is ops midcycle today is intended to build connections, introduce folks to the community and, ideally, generate common feedback to drive change | 16:41 |
mrhillsman | while the ops midcycle in my view is both | 16:41 |
mrhillsman | right | 16:41 |
VW | the spirit of the PTG as it has been has been to allow dev teams to work on specific items for the N+1 release | 16:41 |
VW | so that should be honored too | 16:42 |
VW | in what ever the combined event looks like | 16:42 |
spotz | WHich from the letter would be done in nightly mingling and the possibility of in the hallways | 16:42 |
mrhillsman | i just think one of the key things we are missing from ops midcycle - again my perspective and what i have noticed - is the getting together and work on this piece | 16:43 |
VW | yes - we need to push hard on getting more "tangible" output | 16:44 |
VW | but still give a safe space for new ops/users to get integrated | 16:44 |
spotz | mrhillsman: So say folks interested in X tpoic or issue getting together to figure out what they need or how to do it? | 16:44 |
mrhillsman | i only have one example i can make sense/explain well i think | 16:45 |
mrhillsman | monitoring | 16:45 |
mrhillsman | we have had a session about it at everyone ops midcycle | 16:45 |
mrhillsman | and the session has always ended up being a list of tools folks use at the end of the day | 16:47 |
mrhillsman | i think it would be good for that to move to, most of us use nagios, what can we do to improve the state of nagios+openstack | 16:47 |
*** jamesmcarthur has quit IRC | 16:47 | |
spotz | So would it be more helpful if on a second day or the next midcycle folks do a workshop on how they use the top 3-5? | 16:47 |
mrhillsman | and then when the next midcycle happens | 16:47 |
mrhillsman | still have the what are you using, here's the currently most supported/working, you could start here and help work here | 16:47 |
mrhillsman | https://github.com/search?q=org%3Aopenstack+osops | 16:48 |
mrhillsman | https://github.com/openstack/osops-tools-monitoring | 16:48 |
VW | I don't love monitoring as an example only because I don't think it's "openstack's" place to solve monitoring | 16:48 |
mrhillsman | https://github.com/openstack/osops-tools-monitoring/tree/master/nagios-plugins | 16:48 |
VW | but I understand your point | 16:48 |
mrhillsman | cool, please help with better example lol | 16:48 |
mrhillsman | i just cannot get my brain to do it | 16:49 |
spotz | I cam from a monitoring company so love good monitoring but... :) | 16:49 |
mrhillsman | and i would hate for the point to be missed because of the example/analogy being bad | 16:49 |
mrhillsman | it is a pain point imo like deployment tools | 16:50 |
mrhillsman | we did not have, now we do | 16:50 |
mrhillsman | not that we are going to solve every op issue of course | 16:50 |
spotz | So is that maybe a good example to use, it tells a success story | 16:50 |
mrhillsman | but the general big ticket stuff feels like | 16:50 |
mrhillsman | we should have some collaboration around them | 16:51 |
mrhillsman | and there are devs interested but stop participating when ops are not involved | 16:51 |
mrhillsman | again this has been my experience since first midcycle | 16:51 |
mrhillsman | like right now, i got messages from dev/devops/ops folks who attended tokyo about those repos and ops tools in general | 16:52 |
spotz | example? Trying to think through the whole isssues to help word | 16:52 |
mrhillsman | sorry spotz can you elaborate | 16:58 |
spotz | So in regards to those repos, what are the messages you're getting? | 16:59 |
mrhillsman | well they are inactive primarily | 16:59 |
mrhillsman | and what happens is activity increases because of visibility from the midcycle but it dies down | 17:00 |
spotz | mrhillsman: But ops are activilely using or they would use if supported? | 17:00 |
mrhillsman | they would use if active | 17:00 |
mrhillsman | some are indeed using | 17:00 |
mrhillsman | https://review.openstack.org/#/c/536992/ | 17:00 |
mrhillsman | https://review.openstack.org/#/c/524144/ | 17:00 |
mrhillsman | some community folks who have been around use | 17:01 |
mrhillsman | and new folks gravitate to them because of course they can potentially help them get up to speed quickly | 17:01 |
mrhillsman | with PoCs and the like | 17:01 |
spotz | And are we sure the projects don't have their own? I know OSA does. | 17:01 |
mrhillsman | yeah, i was part of the push for that with osa | 17:02 |
mrhillsman | because of the lack upstream | 17:02 |
mrhillsman | from my time with osic | 17:02 |
mrhillsman | we wanted to do it upstream | 17:02 |
spotz | Just trying to figure out if devs don't know they should be helping there or if ops don't know they exist elsewhere. AKA major communication failures | 17:02 |
mrhillsman | non-tool specific rather | 17:03 |
mrhillsman | from the guy who started the repos, he said he felt like not enough folks were getting involved | 17:03 |
mrhillsman | sorry, lost connection for a moment | 17:04 |
mrhillsman | we're going to start back having meetings next month in an attempt to try again | 17:05 |
spotz | Just trying to figure out if devs don't know they should be helping there or if ops don't know they exist elsewhere. AKA major communication failures -repost in case missed | 17:05 |
mrhillsman | and i am hoping with openlab resources we have a chance to have an environment where these things can be tested | 17:05 |
mrhillsman | yeah, comms is a part of the issue | 17:05 |
mrhillsman | but i think even with comms, not having a workflow that is solid may be the root | 17:06 |
mrhillsman | imho the repos have been serving as catch-alls | 17:06 |
mrhillsman | that is getting into the weeds a bit of course | 17:08 |
spotz | Well a few things just looking about, 'a location for OPerators' then a link to 'developers.html' I don't think I've ever had an OPs person show up at the git and gerrit lunch and learns | 17:08 |
spotz | Oh and use at your own risk:) | 17:08 |
spotz | But yeah this next repo I looked at is 2 years old:( | 17:10 |
spotz | nova has something 4 months old | 17:10 |
mrhillsman | yeah it's spotty | 17:11 |
spotz | So maybe a great forum tpoic would be on those repos, how to contribute, how to ask a project for help? | 17:12 |
mrhillsman | i think first would be for osops | 17:13 |
spotz | I'm assuming it's ops maintained for the most part with dev assistance? | 17:13 |
mrhillsman | yep | 17:13 |
mrhillsman | here is that project for osa - https://github.com/openstack/monitorstack | 17:14 |
mrhillsman | it came out of a lack of collaboration in osops around monitoring | 17:14 |
mrhillsman | unless it went a different direction, it is not specific to osa right, but it is maintained by the osa folks well | 17:15 |
spotz | Not many folks ask for help on it in the channel, this is the one I was thinking of https://github.com/openstack/openstack-ansible-ops | 17:15 |
mrhillsman | ah yes, that one too :) | 17:16 |
mrhillsman | the osic devops team i was working with was helping with that | 17:16 |
spotz | Yeah it's still maintained when need be:) | 17:17 |
mrhillsman | for example the influxdb stuff we heavily worked on and probably why it is not touched as much | 17:17 |
spotz | But OSA is heavy ops and has great relationships both ways | 17:17 |
mrhillsman | because one of the things we wanted to do was use it to build up ideas around doing opsy things with your openstack | 17:17 |
mrhillsman | yes, that is the ideal state i think for all things that sit in the middle | 17:18 |
mrhillsman | so it is a deployment tool which is enough of a concern that it galvanizes folks who contribute and/or use openstack | 17:19 |
mrhillsman | we were trying to push monitoring, fleet management, etc | 17:19 |
spotz | So we can see who else are in the middle and make sure everyone is talking between the 2 sides. But how can we get ops/devs to talk more elsewhere? | 17:19 |
mrhillsman | i do not think we can have a single solution honestly | 17:20 |
mrhillsman | SIGs i think are going down the path to helping | 17:20 |
mrhillsman | co-locating/re-branding/adjusting PTG/Ops Midcycle | 17:21 |
mrhillsman | the Forum | 17:21 |
spotz | Everyone deploys differently, a single overall solution would not be good. But like you mentioned a workflow | 17:21 |
mrhillsman | So i think maybe getting that work^ | 17:21 |
spotz | A how to get help, a how to communicate, etc | 17:21 |
spotz | Even just a point of contact | 17:22 |
mrhillsman | we started working on this right - https://www.openstack.org/community/ | 17:22 |
mrhillsman | and i believe Jimmy, Lauren, etc are always working and wanting feedback - https://etherpad.openstack.org/p/UC-Rocky-PTG | 17:22 |
mrhillsman | line 108 | 17:22 |
mrhillsman | so i think we are moving great down the path of getting folks introduced into the community and over time will understand where we can make additional adjustments | 17:24 |
mrhillsman | we have moved quite a bit from my original post lol i think | 17:26 |
spotz | So I think you have the content of your post:) | 17:26 |
mrhillsman | but yeah, just wanted to get some feedback | 17:26 |
mrhillsman | hehe | 17:26 |
mrhillsman | definitely | 17:26 |
spotz | No I think it's all related | 17:26 |
spotz | That's why I like talking things out and asking questions, you get to the meat of things | 17:26 |
mrhillsman | ++ | 17:27 |
spotz | Wow San Antonio meetup hasn't done anything in a year:( | 17:28 |
mrhillsman | :) | 17:28 |
mrhillsman | houston has not done anything in a year at least too i think | 17:28 |
mrhillsman | well, we participated in the 24 hour devops thing | 17:28 |
mrhillsman | but that was all online | 17:28 |
spotz | I've gone to 2 of the Austin ones | 17:29 |
spotz | helps living in the middle | 17:30 |
mrhillsman | i am thinking about moving nearer to austin, not sure yet | 17:30 |
mrhillsman | but for the activity compared to Dallas, Houston, San Antonio | 17:30 |
spotz | I like Austin, not necessarily the traffic or the cost:) | 17:31 |
mrhillsman | just unreasonably far for "water cooler" talking/networking | 17:31 |
spotz | I miss that part of Castle for sure | 17:31 |
spotz | Ok I just joined SA and Austin from our groups though already in the meetup for Austin | 17:34 |
mrhillsman | ++ | 17:35 |
mrhillsman | actually just reminded me of something | 17:35 |
mrhillsman | need to email ashlee to email ambassadors to ask their user groups for forum topics :) | 17:37 |
spotz | Yep:) | 17:39 |
*** jamesmcarthur has joined #openstack-uc | 18:12 | |
jamesmcarthur | just catching up on this | 18:16 |
jamesmcarthur | Based on teh discussion today, it still feels like Ops Meetups are stuck on either "nobody is engaging, so we have to present" or "people are upset about X" | 18:17 |
jamesmcarthur | It feels like if we can provide a framework for conversation and how to improve the software, that would be very well received. | 18:17 |
jamesmcarthur | And result in good engagement. | 18:17 |
mrhillsman | i think from the improving software perspective we have the things in place | 18:29 |
mrhillsman | and are moving in a good direction | 18:29 |
mrhillsman | software being "openstack proper" (nova, neutron, glance, etc), not tooling related | 18:30 |
mrhillsman | but there are things that ease using/operating openstack | 18:30 |
mrhillsman | that are relevant primarily to ops midcycle | 18:31 |
mrhillsman | and so with the proposed ptg/midcycle changes, how do we ensure that work is done, as imho it is the missing piece coming out of midcycle | 18:32 |
mrhillsman | i am not sure deployment tools came out of midcycle but are a good example of bringing home the point | 18:32 |
mrhillsman | rather than spending a large number of cycles with, "openstack is too hard to deploy", x thing is hard for users/operators/app-devs, we can work on y thing to address x | 18:34 |
mrhillsman | vs x thing is hard discussion being prolonged | 18:34 |
jamesmcarthur | Right - I think you have to set goals for the attendees. | 18:35 |
jamesmcarthur | Like the first thing that should be said at the start of the Meetup is: Our goal is to identify 5 ways in which we can make the software better for users. | 18:36 |
jamesmcarthur | Or somesuch | 18:36 |
jamesmcarthur | And just have every work session focus back on those primary goals. | 18:36 |
mrhillsman | yes | 18:36 |
mrhillsman | also in addressing chris' concern which is related | 18:36 |
mrhillsman | if we are talking about combining | 18:37 |
mrhillsman | going back to matt's what vs how we do that more or less at the ops midcycle granted varying degrees | 18:38 |
mrhillsman | "then we need to honor the intent that a large part of what is ops midcycle today is intended to build connections, introduce folks to the community and, ideally, generate common feedback to drive change" | 18:41 |
spotz | I don't see where the 'intent' isn't kept and perhaps made more beneficial as long as we make sure neither side seems slighted | 18:53 |
* VW needs to read lots of scroll back | 18:57 | |
*** jamesmcarthur has quit IRC | 19:11 | |
*** jamesmcarthur has joined #openstack-uc | 19:13 | |
mrhillsman | people get caught up - rightfully so at times - on words | 19:18 |
mrhillsman | so i just wanted to be sure we thought a bit more about the messaging and had very clear messaging | 19:18 |
mrhillsman | and also again from a personal opinion/hope/perspective to think about work beyond the projects used to build "openstack" to those that help use/operator | 19:19 |
mrhillsman | operate | 19:19 |
*** VW_ has joined #openstack-uc | 19:21 | |
*** VW has quit IRC | 19:25 | |
*** VW_ has quit IRC | 19:25 | |
*** VW has joined #openstack-uc | 19:31 | |
*** ChanServ sets mode: +o VW | 19:31 | |
*** jamesmcarthur has quit IRC | 20:55 | |
*** jamesmcarthur has joined #openstack-uc | 21:03 | |
*** martial has joined #openstack-uc | 21:04 | |
*** jamesmcarthur has quit IRC | 21:04 | |
*** jamesmcarthur has joined #openstack-uc | 21:04 | |
*** VW has quit IRC | 21:16 | |
*** martial has quit IRC | 21:35 | |
*** jamesmcarthur has quit IRC | 21:39 | |
*** VW has joined #openstack-uc | 21:41 | |
*** ChanServ sets mode: +o VW | 21:41 | |
*** VW has quit IRC | 21:51 | |
*** jamesmcarthur has joined #openstack-uc | 22:18 | |
*** jamesmcarthur has quit IRC | 22:41 | |
*** jamesmcarthur has joined #openstack-uc | 23:24 | |
*** jamesmcarthur has quit IRC | 23:29 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!