I wrote this on the plane on the way to the PTG and then planned to write one each day. It turns out the PTG was (for me) extremely busy and very exhausting so not only did I not get around to writing a posting each day, I also didn't get around to posting this until two weeks later. I still hope to write a reflection on the week as it actually happened but this provides a good view on what I thought would happen.
This coming week is the Project Team Gathering (PTG) for OpenStack. A week where contributors to the many OpenStack projects gather to discuss the details of upcoming work and distil plans. Initially I was not going to be able to attend; my employer had not allocated funding. This is despite the fact that my role with the company is to work effectively 100% upstream (within the open source community rather than on internal projects). Thus I had made no plans to come.
With encouragement from others I applied for and received a very generous travel grant from the OpenStack Foundation to attend the PTG for the full five days. I'm writing this from a plane over the Atlantic.
The PTG is divided up into two sections. The first two days are devoted to what's called "cross project" work. This is work that is relevant to all the various projects (Nova, Heat, Kolla, Telemetry, etc.) that make up the OpenStack universe; changes that need to happen across all of those projects, or discussions to resolve issues of identity and direction relevant to everyone.
I'm a member of the API working group. We have a room on Monday and are sharing a room with the architecture working group on Tuesday. We intend to talk about capabilities discovery (how an end user is able to discover what's "turned on" (or off) in this cloud, consistent service catalogs and unversioned endpoints in the service catalog, and guidelines for the ways in which changes in an HTTP API must signal a change in the version of the API.
That's a trend. All of these things impact interoperability amongst multiple clouds. Something that is high in everyone's priorities.
The latter three days are devoted to discussions within individual projects. I'll be hanging out with the Nova (compute) project. The list of topics is huge but most of my attention will be on those related to resource providers and the placement HTTP API. These data structures and APIs are for keeping track of qualitative and quantitative resources (such as RAM or SSD disks) used for making placement (scheduling) decisions when allocating resources to user requests.
This will be an interesting week: in the past six months OpenStack has gone through a contraction. The number of humans that are available to devote themselves to upstream work has shrunk. The number of companies that are making direct investment in the improvement of OpenStack has shrunk. Many projects, certainly Nova, have been overstretched for a long time and now the situation is even worse.
As we have our discussions this week a lot of great ideas will need to be dismissed because there simply won't be the people to do the work and many people (myself included) will struggle to commit to doing work because their status as employees who do OpenStack is either definitely coming to an end or is at least at risk.
Despite all that, I'm looking forward to the week. It is an excellent time to collaboratively share and build ideas and catch up with colleagues. More tomorrow after the first day.