Third time's a charm, right?
Well, yes and no. Overall I would call the summit a success, and a lot of good things came out of it. It wasn't all sunshine and puppies though, and there are definitely some things I personally, and we as a community, could improve. Which means I should have lots of interesting things to talk about in this post. Let's get right to it.
This was an interesting summit for me personally. Due to the fact that TripleO started shipping as a tech preview in RHEL-OSP 6, and will become an increasingly important part of future releases, I've had a lot more downstream responsibilities than I had in past cycles. It's been a new challenge to balance upstream and downstream work, and I'm still not sure I've struck the right balance. I suspect this will be an ongoing source of tension no matter how much experience I gain, however.
Perhaps the best example of the shift in my responsibilities was that for the first time I was invited to a customer meeting. It turned out I didn't need to be there at all and contributed nothing, but the fact that someone even thought I needed to be there is a big change. Hopefully in the future I can learn to navigate these situations better and ask the right questions to determine whether such meetings are going to be valuable or a waste of my time (while recognizing that those two things are not mutually exclusive ;-).
I can't very well go any further without giving a big shout-out to the location. Vancouver was very nice, and the conference center itself was downright spectacular. One of the highlights of the week was definitely holding a couple of Oslo work sessions on a terrace overlooking the water. Getting interrupted every five minutes when a float plane took off was not ideal, but totally worth it. The attached picture was taken during one of these sessions.
Speaking of Oslo, we have a mascot now! It's the moose, and if you saw some weird guy walking around the summit with a stuffed one wrapped around his neck, that was me. We were also handing out moose stickers all week to spread the Oslo love, and I'm told it got carved into a totem pole somewhere. Unfortunately I never did get a chance to see it myself. All in all it was a fun way to increase awareness of Oslo, so we might have to do something similar for TripleO in Tokyo.
First, a mea culpa: I was under-prepared for some of the sessions I was leading. I won't go into why this happened since it would just be a bunch of lame excuses/justification (depending on how generous you want to be), but it was not ideal. I also missed a few sessions that I would have liked to have attended, due to conflicts with TripleO sessions I had to attend.
Which brings me to the first not-so-great thing from this summit: we had too many sessions for Oslo. This may sound like a strange complaint, but having ~17 sessions (may not be the exact number, but it's close) made it very difficult to coordinate my schedule, and I heard from others interested in Oslo that it was a problem for them too. In addition, it allowed us to accept some session proposals that probably didn't require face-to-face discussion, which meant we wasted some number of those sessions. And to be clear, I'm not pointing fingers at anyone else here. Some of my sessions were the biggest offenders.
This is an area where we're still learning as a community. Since I started working on Oslo, we've always had a very limited number of sessions available, so we had to be quite careful with our selection process. Because we had an essentially unlimited number this time, we held some that we shouldn't have. The topic was discussed in a recent Oslo meeting, and everyone seemed to be in agreement that even when we can hold a large number of sessions, in the future we need to look more carefully at whether we should. We also don't expect it to be a problem in Tokyo, where session space will likely be more limited again.
I don't want to give the impression that all of the Oslo sessions were a waste of time though, because that was definitely not true. There were a couple of sessions about oslo.versionedobjects, which is a big piece of the rolling upgrade process that needs to be integrated into more projects this cycle.
Despite lacking a few key people, the session on alternative backends for oslo.config was productive, with input coming from a few new sources (so maybe it was okay that the conversation wasn't dominated by people who had already been involved? Food for thought...). I also felt the oslo.messaging sessions were excellent, with some new conclusions reached and, as always, some new questions raised that require more follow-up. In any case, it was good progress on a critical piece of the OpenStack ecosystem.
I was going to talk about TripleO in this post as well, but I think it's long enough as-is. I'll save TripleO for a separate post. Stay tuned!