Clever Geek Handbook

How I didn't go to London, but participated in the London DevOps Enterprise Summit





The regular DevOps Enterprise Summit was held in London from 23 to 25 June 2020. “In London” can be written in quotation marks, as the pandemic did its job and the conference was held online. It has both bad things here (networking still suffers a lot) and good things: you can take a break between reports away from all people, you can set aside a whole few days to study the reports that interest you, and focus - this is very useful ... Just recently I was at a meetup "in Chelyabinsk", the next day - "in Novosibirsk", and a few days later - in California: physically it would be extremely difficult to do this. Of course, there are recordings, but the very sense of making time for learning seems to help learn.



The formats of conference reports have already taken root on Habré, and the desire to develop in conditions when you sit at home has become especially strong (you have to do something) - and here the question arises: there are a lot of conferences ... how to see everything that you want to see?



And here I thought that I could try to make a report on the conference in a different format - not to tell about what the reports were (for this I posted a separate "live text broadcast" from those reports that I listened to), but to write an article that summarizes conclusions and current trends, will indicate where to turn for the latest knowledge.





What is the DevOps Enterprise Summit? How is it different from other conferences?



Any article about DevOps conference, of course, should start with a discussion of what DevOps is - and, of course, everyone is tired of it. Therefore, I will try to be concise!



In fact, everyone who argues is likely to agree on at least one thing - that "devops is a set of practices that facilitate software delivery and infrastructure management." The only differences are that one group (and most often these are “admins”) says “discussion of practices should be practical”, and the second group (and this is development and, most often, management in development) believes that “this is philosophy and methodology ".



I would risk incurring anger for "simplification", but in essence: for development, the introduction of "admin" methods and capabilities changed the development process itself ideologically and turned DevOps into a new Agile - when we hear about management conferences about DevOps, we hear about changes in methodology development. And the DevOps Enterprise Summit is one of those conferences that are useful for "admins": there were also reports closer to technical.



So, the DevOps Enterprise Summit has been held since 2014, and it is organized by IT Revolution with the founder of Gene Kim, known to many from the book "Project Phoenix". IT Revolution is the publisher that has launched a series of books on DevOps and its links to Agile. Why Enterprise? This is a particularly important point.



Implementing agile methodologies in a small company is easy: in multi-person companies, they most likely have existed from the beginning - you just need to call them that way. With companies of thousands, tens of thousands of people, everything is much more complicated. There are a huge number of processes in them, built so that this colossus continues to move, and this can be compared to a large ship. When a ship travels across the ocean from one mainland to another, it fulfills its role, but the ship cannot change the route and change course in the shortest possible time - this takes time, especially in technology businesses. A young company of just a few can build a prototype in a matter of days that a company of thousands will build for years - if it doesn't change.Therefore, large companies want to change and use the methods of the "young", but they need to understand how not to disrupt their processes. I have already given an example in another article: in Excel 1900 is considered a leap year, and this is done for backward compatibility with versions of Lotus 1-2-3, which had such a bug. These versions were released in the 80s - and the question is: should this compatibility be maintained? To what extent can bugs be allowed in the “we will fix it later” approach when you install software in banks on servers that are generally disconnected from the Internet? How can you maintain that same flexibility? This was what the conference was about.These versions were released in the 80s - and the question is: should this compatibility be maintained? To what extent can bugs be allowed in the “we will fix it later” approach when you install software in banks on servers that are generally disconnected from the Internet? How can you maintain that same flexibility? This was what the conference was about.These versions were released in the 80s - and the question is: should this compatibility be maintained? To what extent can bugs be allowed in the “we will fix it later” approach when you install software in banks on servers that are generally disconnected from the Internet? How can you maintain that same flexibility? This was what the conference was about.



Format



Every conference tries to figure out how to get itself online, and it's interesting. In DOES, the following can be noted:



  • The reports are still split by track, the viewer can switch between tracks in the process.
  • The reports were recorded before the day of the conference, in order to avoid technical problems with the broadcast, the speaker during the report is in the slack channel and communicates with the audience. This is probably the most interesting thing that I learned from the organization. The solution is contradictory - on the one hand, the interactive disappears, on the other hand, much more time for questions (and much more opportunities for the speaker to answer more questions). However, for example, I mainly listened to the report and did not want to switch to discussion, so I came at the end when there was little time left and a new speaker came to the next report.
  • , -, .
  • , , , , « », - , , , .






DevOps Dojo. . , . SRE- Uptime community, , ? Dojo — , , . DevOps- Target, , , , , , , . . youtu.be/1FMktLCYukQ
  • Swiss Re — , 156 , DevOps- . , « ». «» — «» CEO, , .. , , , Gartner-, 76% DevOps- «, ». . — , . , «» DevOps- — . . , , : Hermes Germany GmbH , CIO, « », «» .
  • , — . State of Devops, Gene Kim Accelerate.
  • , , DevOps And Modernization 2.0 (CSG) Scott Prugh — DevOps . , — , : Cobol-, 3,7 IBM High Level Assembler Java ! : .


Mentioned books worth reading:

Team Topologies in Action , Accelerate: Building and Scaling High-Performing Technology Organizations .



And I am a little more regular than the blog here, I run my own telegram channel , subscribe.


More articles:


All Articles