Kubecon was held in Detroit last week and we thought it would be interesting to recap some of the events from the Tracetest team’s perspective with an eye towards Observability. Tracetest had a booth on the main convention floor, but we wanted to attend and participate in a couple of OpenTelemetry specific events on the Monday & Tuesday before Kubecon really kicked off, so both Sebastian and I arrived on Sunday. We were joined in Detroit later in the week by members of the other Kubeshop product teams - Monokle, TestKube, Kusk, KubeFirst & BotKube. Kubeshop (our mother company) is 100% remote, so it was fun to have time to hang out with other members of the company.
First, the venue. We stayed in downtown Detroit and we were impressed by the revitalization of the area. Great restaurants and bars provided lots of entertainment ‘after hours’. The conference center itself was beautiful, with huge views looking out on the Detroit River.
## Day 1 - Open Observability Day
Monday we attended the Open Observability Day at the convention site. Did the onsite check in for Kubecon, got our badges, and found the conference room. Three rooms were combined to form one large room for the presentations - I would estimate attendance at 100+.
The first keynote really resonated with me by Ian Smith titled ‘The Struggle is Real’. It looked at the state of adoption of OpenTelemetry and discussed how to add more value. It described the current state of distributed tracing thus: “Tracing has become the high-promise, high-effort, low value story”, with tools primarily being used by just a few power users. Our mission at Tracetest is to change this. We believe by basing end-to-end testing on the data in a distributed trace utilizing trace-based testing, tracing will become important across multiple roles, with developers, QA/Testers, and SREs all obtaining value from the investment.
There were many talks throughout the day - several focusing on the OTel collector (including OpAmp!), one by Vijay Samuel showing eBay’s journey, and a session near the end by Libby Meren discussing how to build a business case for implementing observability in your organization.
## Day 2 - OTel Unplugged
The OTel Unplugged event was held off site about a mile south of the convention center in a beautiful older building which had an area converted to an event center. The venue was much smaller and intimate than Monday’s, leading to many more ad hoc conversations and meeting new and old friends. I would estimate the attendance at around 75 people. There were many maintainers, vendors interested in the project, and end users of the OpenTelemetry project.
I found the sessions very interesting. The day began with a ‘State of OpenTelemetry’ discussion. There were several presentations - highlights for me were the maintainer and end user panels in which questions were asked of the panel by attendees and people joining the event remotely. After a delicious lunch, there were a couple of breakout sessions based on topics which were voted for. I participated in the Real User Monitoring (RUM) and OTel user experience sessions. The meeting ended with adult refreshments and great conversations. The event was VERY WELL run and organized - you could tell a lot of effort was put into preparing it.
## Day 3, 4 and 5 - Kubecon Convention Floor
Day 3 saw Kubecon kicking off in high gear. Sebastian and I both worked the Tracetest booth, so my impressions will be based on the floor and not sessions or keynotes. The room the vendor booths were set up in was cavernous - it was simply huge. The major commercial vendors had booths equal to the scale of the room, they were quite elaborate and had large footprints. I saw the amount of work Kelly and Tina at Kubeshop went through to prepare our team for Kubecon, and it was obvious that this level or preparation was widespread. It is truly amazing to think of the 1000s of hours that go into planning and preparing for Kubecon.
Observability was front and center throughout the floor, with all the major commercial vendors present and showing their observability solutions. Datadog, NewRelic, Lightstep & Dynatrace all had large booths with big footprints. There were also a number of smaller vendors present. The OpenTelemetry project had a dedicated booth that was manned by volunteers which had very good traffic each time I walked by.
Traffic at the Tracetest booth was steady but not never overwhelming, with the quality of the conversations being extremely high. This sentiment was reflected by others we spoke to on the floor. Medium flow of people, but there were not a lot of people just coming by for the t-shirts. The attendees were truly interested in the Tracetest project and wanted to learn more about how it could help their company. Both Sebastian and I were kept busy showing how Tracetest brings trace-based testing to your current distributed tracing environment. We must be doing something right both with the tool and at the show - one of the curated list of impressive standouts from Kubecon listed Tracetest as ‘best in show’:
“My ultimate pick was Tracetest which was a phenomenal show stopper for E2E testing and tracing use cases. A phenomenal improvement over other tools in the space that I’ve seen.”
In summary, we left Kubecon and Detroit excited about Observability and our open source project. The OpenTelemetry community is simply the best, as everybody is friendly and motivated to build an open standard that can be adopted successfully by both end users and vendors. We had several of the key contributors come by and see a demo of Tracetest and offer their ideas and thoughts for direction, and it was much appreciated.
Interested in Tracetest? You can install it in 5 minutes, connect it to your existing tracing infrastructure, and build your first test. Once installed, let us know what you think! You can add issues on Github or join us on our Discord channel to discuss the future of Trace-Based Testing with Tracetest!
This blog post documents our findings and provides step-by-step instructions to set up a local observability environment and inspect traces on its API, node agent, and container runtime.
Trace-based testing overcomes the limits of black box testing by leveraging work already invested in observability of your system via distributed tracing.
Join the Tracetest Community in one of these channels: