With twins on the way (due end of July) I'm planning on taking 6 weeks off, so I won't be able to commit any time to this before October at the earliest.
If those timelines still work, I can contribute some chapters on HA and scaling/tuning your Airflow cluster (going in to detail in what the different tuneables around scheduler behaviour mean and when you might want to change them) (We should also make sure that we focus on the changes to the user and why they might care/what they have to do -- the internals of how we did a change probably isn't that interesting?) The other thing to bear in mind: how much time are Packt going to invest in this? i.e. how good an editor do they have? I have heard some horror stories from authors writing for Packt in the past. https://www.reddit.com/r/humblebundles/comments/83x1l4/warning_packt_books_suck/ is one of the such public cases that I can find quickly, but I do have more I can't share. Caveat author basically. -a On Jun 1 2020, at 12:25 pm, Jarek Potiuk <jarek.pot...@polidea.com> wrote: > Hello Everyone, > > I have been approached by Packt Publishing <https://www.packtpub.com/> > about writing a book about Apache Airflow. I discussed it with a few people > involved in Airflow development (including Bas as he is finishing the > Manning Book). > > I thought this might be a great idea not to repeat what Bas wrote, but > maybe we could write a book focusing on new Airflow 2.0 features - > especially that expected publication time is around 5-6 months from now. > Which will be appropriate timing for Airflow 2.0 release to settle and > possibly to start getting traction. And I think it might be one of the > important factors in faster adoption of Airflow 2.0 > > The idea would be to write about: > - functional dag writing > - high availability > - migrating to Airflow 2.0 (including backport packages) > - airflow 2.0 APIs > - autoscalability, using prod image, helm chart etc. > - .... other :) > > I spoke to Bas and he told me those things (most of it I kind-of expected): > > - writing a book is a hard one > - direct profits are nowhere near the money you can get in your day job > - it takes much more time than you expect > - the biggest benefit you can get is to have your name on it. > - Bas is perfectly ok with other people writing the book and thinks > it's a > good idea ("the more the better!") > > I think writing such book by one person is quite an effort indeed, but > - as > we already saw from the "What's coming in Airflow 2.0" NYC meetup > presentation - if we involve more people who work in concert (or I'd say > orchestration ;) ) that might be much more doable. Especially that Packt > offered coordination of the efforts with all the people who'd participate. > > So I wanted to see if there is enough interested people who are active in > the Airflow community and would like to participate in writing the > book. Of > course, that means that we would have to all share the profits (but I am > perfectly OK). > > If you are at all interested - please head to a new channel I created in > slack https://apache-airflow.slack.com/archives/C014SM7QB1A > ("book_airflow-2_0"). I would like to discuss there potential chapters and > people who would like to take on writing those. > > Looking forward to getting the involvement of the active community members! > > J. > > > > > -- > > Jarek Potiuk > Polidea <https://www.polidea.com/> | Principal Software Engineer > > M: +48 660 796 129 <+48660796129> > [image: Polidea] <https://www.polidea.com/> >
