GeekSocket Plug in and be Geekified

Rootconf Pune 2019

Since 2017’s Rootconf, I always wanted to attend the conference. But due to college exams or office work, I wasn’t able to attend any of the Rootconf till date. Rootconf is the conference which covers topics like DevOps, site reliability, distributed systems, security and infrastructure at scale. These are the things which I find really interesting and that’s why I always wanted to attend the conference.

Few months ago, Chandan told me that, Rootconf will be coming to Pune soon. This means there will be Rootconf Pune edition along with the Rootconf which happens every year. I was really excited about this. I also wanted to volunteer for the conference.

Submitting talk for the conference

Over the past one year, while working with InfraCloud, I had few ideas for a talk. One idea was Tracing Cassandra with Jaeger and the other one was about implementing distributed tracing in FaaS. Thanks to Vishal and Anwesha, who pushed me to submit a talk proposal for Rootconf Pune.

With a few follow up questions, the talk proposal got shortlisted luckily and then it got selected. So the next task was the rehearsals. I really liked the idea of having rehearsals for talks, HasGeek folks do this for each conference, which is really awesome. Conducting rehearsals requires good amount efforts.

During the first rehearsal, I got really constructive feedback. Most of the things were exactly the same which I had thought. Those were the suggestions which helped me to shape the talk as same as per my expectations. Special thanks to Talina, Anwesha, Zainab, Aaditya, and Cyrus for giving suggestions. Vivek also helped me to do the dry run of the talk and he also reviewed my slides before each rehearsal.

With all the changes and improvements, I was ready to deliver the talk.

The Prometheus workshop

One day before the conference i.e. 20th September, there was Prometheus workshop by Goutham. I attended it with a few of my colleagues. He covered almost all the things related to Prometheus and it’s components. He covered PromQL basics, which were helpful for anyone who is starting with Prometheus.

The conference day

On the day of conference, I reached the venue around 07:30 with Chandan, Anwesha and Gaurav. I was volunteering as well. After setting up the stage, camera, projector etc we were ready to start the conference. It was a single track conference. I was doing work of mic runner. My job was to pass the mic to person who wants to ask questions to speakers. I attended all the talks in the first half. Check out the schedule for the list of talks and the proposals for them.

In the second half, there were BOF sessions, where people discuss about a specific topic. I attended ‘Technology Selection in Cloud Native Era‘ which was after my talk. We had a good discussion about what all tools people are using and on what basis they selected those tools during that BOF.

Using DNS as a layer of defense’ was an enlightening talk by Swapneel. We also had discussion about DNS over HTTPS during Q&A of his talk. Do check out the video of his talk over here.

The last keynote by Piyush Verma was really great. He covered details about site reliability of a distributed systems and what are the factors we should consider while building a distributed system. You can checkout the video of his talk here.

Rootconf Pune was the first among the separate editions of annual Rootconf to be conducted. It was successful, thanks to a great combination of co-ordination, planning and execution.

My talk on distributed tracing in FaaS

This was my first conference talk ever. Rootconf gave me the opportunity to talk about something which I worked over the last year. The talk was about how an implementation of distributed tracing should look like. It covered basic details about the term observability and the need to distributed tracing. I also talked about function as a service, microservices, Fission and Jaeger.

Though I have delivered multiple talks in meetups, I was a little bit nervous at first but I became comfortable after 5 to 10 minutes. Also I had to use Amol’s laptop as my laptop wasn’t even detecting the projector. I had my demos recorded, so I just copied them on his laptop and I was good to go.

During both the rehearsals I finished exactly in 35 to 37 minutes. There were many concepts as well as slides, still I was able to cover everything in exactly 38 minutes with just 2 minutes for Q&A. Few people asked questions about overhead of tracing instrumentation done in the applications. Also 2 to 3 people reached out to me after the talk. We had a nice conversation about the engineering efforts required to implement tracing and how those can be minimized. It was a great experience all together.

References


Comments

comments powered by Disqus