I’m just back from attending the annual DevOpsDays Edinburgh 2019 Conference. DevOpsDays is a distributed set of local tech conferences held in multiple cities all over the world.
DevOps is a bringing together of the traditional “Dev” i.e. programmer/developer and “Ops” i.e. systems/network/storage administration disciplines of traditional IT teams. It attempts to break down the silos between these 2 camps and allow a more efficient and functional approach to software development and management.
DevOps broadly adopts the Agile approach and as well as promoting closer interaction between the “dev” and “ops” roles champions automated building and testing and a release-fast, release-often approach to release management which can seem at odds with how traditional companies manage their technology. It also seems to clash with the strict hierarchy and management structure of traditional companies in favour of looser team groupings such as “tribes” and horizontal product/enablement teams.
The structure of the conference was the main 30-minute talks in the morning, followed by a set of “Ignite” lightning talks lasting 5 minutes each then a set of interactive then finishing up with 30-minute “Open Space” sessions where groups discuss a given topic amongst themselves with an open-door policy.
While this is a tech conference several of the talks revolved around structure and strategy. In essence how companies migrated from a traditional structure, why they did it and their experiences.
There were also talks about complexity theory (analysing the Chernobyl disaster) and Data Science.
The talk about complexity also raised a cautionary tale about the dangers of how a lack of diversity can result in groupthink and missing/misinterpreting events that could have prevented the disaster.
We were also introduced to the “4 V’s” of Big Data: Volume, Variety, Velocity and Veracity and these were used to explain whether the data Alan Turing used to crack the Enigma codes could be classed as being “Big Data” (spoiler- it was!)
The non-tech talk which resonates the most however was titled “Don’t be a hero” and it described the “heroic” behaviours that many technologists show but how these are often at the detriment of their own wellbeing. In my opinion this was a very brave, personal discussion of the speaker’s own burnout and both how he identified it and is recovering from it.
The technology talks had an emphasis on Cloud-Native technologies such as Kubernetes and Docker. This is understandable as many of the companies at the forefront of the DevOps movement are technology companies who make great use of cloud platforms such as Amazon AWS and Microsoft Azure. It is interesting thinking how these tools and practices could be adopted by non cloud-native setups.
Finally the Open Space talks covered a wide gamut of topics supplied by and selected by the delegates. The level of discussion was higher than it had been in previous years- in particular the discussion about the compatibility between DevOps and regulated industries was lively and interesting. In saying that I’m not sure there needed to be Open Spaces held on both days.
Overall I thoroughly enjoyed this conference. The 30-minute talk format was a good “sweet spot” in terms of length and the Ignite talks were all entertaining and useful. The standard of speakers was also very high although there was a fair amount of overlap in topics covered.
I’m not sure how well DevOps would work in the University. It would be interesting to try it out to deliver some projects and it would also break down the “Us vs Them” silos between our various teams and divisions.
