Greetings.
Just a note for anyone looking for ansible in epel7.
It’s been retired there because with the release of RHEL 7.4 it’s now
int the rhel-extras channel.
Accordingly, you can get ansible now from rhel extras channel, or CentOS
extras repo.
You can also get ansible rpms now from
http://releases.ansible.com/ansible/rpm/
Note that ansible continues to be available from epel6.
The last day of Flock started even earlier with a team breakfast. The Fedora Engineering team (which I am on) really only gets together once a year at flock, so we gathered up for a team breakfast to catch up on things and see each other in person. Nice breakfast at the Egg and I.
Then back to the conference where folks were showing off what they did or telling people how their sessions went. I think a number of folks had to depart before this, so I am not sure how well it worked, but it was nice to hear how sessions I was unable to attend went.
Next up was the weekly FESCo meeting. Since most of us were at flock, we gathered in the hotel lobby with laptops and had a IRC meeting while all together. The meeting went pretty quickly and then it was time to catch a ride to the airport. The ride was smooth and trouble free until we got close to the airport. Then an accident made the last few miles take an hour or so. I had time to catch a sandwitch and a beer before my flight (did you know Sam Adams makes a nitro vanilla porter?). Flight was long, but uneventfull and then just the hour or so drive home. Finally got home around midnight local time.
Some final thoughts:
Anyhow, I had a great time as always. Many kudos to the flock team and all the attendees! +1 would flock again. 😉
Day 3 started again at 8:30, so nice and early.
First up was Mike McGrath giving the “What does Red Hat want?” talk. He had a very interesting (and likely right on the money) concept: Fedora lives in the space right after the very early adopters of something and before the bulk of people start adopting something. In other words, if we aren’t adopting new things, stablizing them a bit and then moving on, we are not in the right place. We don’t want something too soon, it has to be starting to gain adoption, and we also don’t want something thats too stable and going on to the masses because it’s boring. Excellent food for thought.
Next was my workshop (luckily nearby Mikes talk). The room was pretty packed, which was great. I first went over plans for the rest of this year, next year and beyond as I see them. I’d like to see us get more apps in openshift. A lot of the simple ones shouldn’t be hard at all. Then, I’d like to look at leveraging some cloud providers some more. Since we have openshift and apps in it we could look at migrating some of them to cloud providers and reduce our dependency on our hardware in our main datacenter. This would also allow us to spread out more and handle more traffic and load more easily. I’d like to also work more with other open source community infrastructures: in particular CentOS folks, who we already talked to about running our jenkins instance and setting up better integration between pagure and centosci. We share a lot of needs and there’s no reason we can’t also share resources to meet those needs. I’d also like to get more metrics than we have these days (about bwandwith used, cpu consumed, etc… which moving to a cloud provider will give us). There will of course be lots of things where that doesn’t make sense (like the buildsystem/releng machines), but it should make us a lot more flexable.
After going over that I opened things up to questions and ideas. I was hoping for some more crazy off the wall ones, but we did get a number of good ones. We talked jenkins and CI, we talked about metrics and things that do that, monitoring (if we wanted to move away from our current setup), how we might best make use of modules (perhaps have infra specific ones we use for our apps, which makes us not be tied to a specific Fedora release), making docs or other artifacts via openshift, or even reworking how packages are built to use it. Then we broke up to hack on various things and let other folks have some time back if there were things they needed to do.
Then after a bit of hallway/fixing compose issues time, it was lunch again.
After lunch was a general talk on modularity. I think I had a handle on things before this, but it was good to see the current state and there were some great questions from the audience.
After that it was back to the hallway to discuss things with folks and work on composes and other fires.
There was no planned dinner this evening, so went out and had a relaxing dinner with some folks at a local pasta place.
Day two of flock started too early for non morning people (8:30am), so I was up around 7am to shower, get some breakfast (at the hotel resturant, it was ok) and get over to the conference rooms.
There were a ton of interesting talks in the morning, but we were also having some rawhide and branched compose issues, so I opted to go to the hallway track and work on that and discuss all kinds of things with all kinds of people. It was great to be able to have full bandwith discussions with someone about something.
Had lunch outside on the deck. It was very nice there, but a bit hot and bright for a bunch of nerds. Lunch discussion was all over the place, including new arm laptop (pinebook) and Fedora builders and how various things were made to laptops and support from various vendors.
Then it was time to rush to the EPEL for the future talk from Smooge. The room was standing room only, so perhaps people are realizing how popular and important EPEL is. Smooge presented some proposals for dealing with various problems in EPEL, but I had to go and get a server that crashed back up and running, so I missed the discussion part of this. Hopefully we will see more discussion in EPEL meetings and on the epel-devel list.
Next up was Ricky’s talk about our recent openshift deployment. He went over how we are doing things and the various caveats and issues around any new tech. The slot was only 30min, but he covered things very well.
Next (in the same room) was Jeremy’s talk about the future of fedmsg. I was a bit confused here I admit (sorry Jeremy!) about the scope and nature of changes. Really it was a bunch of things we could (or decide not to) change. I somehow thought of it as all tied together. I am sure this will also have a bunch more discussion soon and we will see what we all want to do here.
I then went back to the hallway to work on compose issues and some urgent tickets.
Then it was time for the evening event: Wackenhammers arcade. It was a pretty amusing old time arcade with beer and food from a food truck. Got some nice visiting in here with the Fedora kernel team and tons of other folks.
Day 1 started out way too early. 7am (which feels like 4am for me!).
After a quick breakfast downstairs, and run by registration I made it to the ballroom just in time for opening announcements then our Glorious Fedora Project Leaders “State of the Fedora” talk. The room was pretty packed actually, and as allways Matt had some pretty charts and graphs.
I found the premise interesting: That Fedora lives on the innovation curve at the very front left. A small bit behind the very bleeding edge, and back to where it starts getting boring. If thats true (and it seems likely), then we always need to be pulling in the very new stuff from one side and integrating it. So the state of things now is a bit on fire, but in a good way.
After that was the group picture (interested to see how it turned out this time) and then everyone running a session got a minute (or less) to pitch their session and try and get folks interested in going to them.
Then, on to sessions:
Then time for games night + dinner (pizza) and beer. The evening ended with us hanging out in the hotel bar and chatting on all sort of topics.
Loading ...