Last year, we made a big mistake with the DevOps section hall, and gave it the smallest room for 30 people. At the reports, the crowd stood along the walls, at the doors, and even behind them. At the same time, the reports of the section received very high marks. We learned our lesson: devopsers, you will have a great spacious room in the new Congress Hall at the anniversary DUMP.
See under the cut what topics took off last year in Yekaterinburg and Kazan, and what the program committee is waiting for this year
Topics 2019 that "came"
At DUMP Yekaterinburg in April last year, all 5 topics received high marks (above 4,2 out of 5). The theme from Vladimir Lila, the elastic man from Kontur, became the leader. The report is called "Elastic weighing in a petabyte", although by now this threshold has long been left behind by Kontur.
Hear about the organization of the process, the transport of logs, and the technical details of building such a cluster, common mistakes and the benefits of all this:
Victor Eremchenko was the second according to estimates. His topic is "How we reduced rollbacks of server releases by 99%." Victor talked about how Miro approached the process of continuous delivery, and how these approaches helped to reduce the number of rollbacks of the server release; about how it helps teams quickly and conveniently deliver their functionality to production.
The report contains, among other things, real examples of the use of various tools and technical details of the CI / CD process.
On the , which took place in November 2019, for some reason the topics about interaction within the team and between the development and operation teams were well laid out.
The report of Alexey Kirpichnikov (Kontur) βThe Curse of the Infrastructure Teamβ was not registered for technical reasons. Maybe the word βcurseβ played a roleβ¦ But since Aleksey gave this talk on DevOops, we found a link to the post
The theme of Marat Kinyabulatov (SkuVault) βIn the middle of the ashes: post-mortems as a tool for continuous improvementβ also sounds dramatic. Marat talked about post-mortem as a tool (and procedure) for inspection and adaptation. On how it helps teams prevent incidents in the future, demonstrates to management the steps taken, creates a safe atmosphere, giving employees room to improve processes:
The DevOps section at DUMP 2020 is run by 4 program directors: Alexander Tarasov (ANNA MONEY), Konstantin Makarychev (Provectus), Viktor Eremchenko (Miro (ex RealTimeBoard) and Mikhail Tsykarev (ICL Services) who formulated the concept of the section this year.
The concept and topics of the DevOps section
This year I would like to get a maximum of practical solutions, a minimum of theory. Tell us where you were hurt, but it became good. What worked and what didn't. Let's rejoice / cry together.
Here is a list of topics that seem relevant to us for the DevOps realities of 2020:
CI / CD
- Awesome CI/CD pipelines
- GitHub Actions (no theory, just practice)
Cloud
- CI/CD in the Cloud (Spinnaker and others)
- Deep dive into GKE, Kubernetes, Istio, Helm, etc.
- Data in the Cloud (PVC, DB and others)
- Clouds for ML
- Serverless (practice only)
- Clouds in Russia
DevOps/SRE
- How to observe the system (observability): service mesh, monitoring and auditing
- Security (DevSecOps)
- Configuration Management (Ansible, Terraform, etc.)
- Let's talk about culture (Best Practices)
- Shift the Enterprise Stories
- Management: life hacks, useful tips, fuck-ups.
If you didn't find a theme in the list, but you have something to share with the devops community, it's all the same . We will definitely look into it!
Time for the report 35 minutes + 5 minutes questions in the hall. After that, you can communicate with the participants in the expert zone for the entire break of 20-30 minutes.
π
Source: habr.com