Hackathon Day Four: Thursday, June 5th
Tracer's beta launch is nearly here - today we shipped onboarding, fallback tracing, branch-based installs, and better OOM insights.
const metadata = ;
const meta =
Work-work-work! Today was all about the final sprints toward getting our Beta version live and making sure every piece of the Tracer experience is ready to demo. The team spent the day heads-down, testing everything from onboarding to local pipeline runs.
By midnight, our full Tracer demo will be up and running. Here’s what we shipped today to get us there:
1. New Onboarding Page Live on Sandbox
Because of Laura’s efforts last night, the team woke up with a brand new sandbox onboarding page live!
It only took 30 changed files, 6,445 insertions and 21,169 deletions! Our hard work is becoming more visual now, thanks to our new kick-ass front-end page!


2. eBPF Fallback
The Tracer client now falls back to traditional system polling when eBPF isn’t available:
- ✅ Works inside Docker containers (even without sudo)
- ✅ Supports both ARM and x86 EC2 instances
- ✅ Preserves full observability during fallback mode
This makes tracing more robust in constrained environments and ensures Tracer runs smoothly even when kernel features are unavailable.
Thank you David!!
3. Branch-Specific Binary Installation
To simplify testing and avoid overwriting shared binaries, we’ve added support for branch-based installs via our installation script.
Latest dev release:
`bash
curl -sSL https://install.tracer.cloud/installation-script-development.sh | bash -s dev && source ~/.bashrc
`
Specific feature branch:
`bash
curl -sSL https://install.tracer.cloud/installation-script-development.sh | bash -s feature/branch-name && source ~/.bashrc
`
Latest stable release:
`bash
curl -sSL https://install.tracer.cloud/installation-script-development.sh | bash && source ~/.bashrc
`
4. Better Debugging with Out of Memory Visibility
We’ve improved how Out of Memory (OOM) errors are tracked in Grafana:
- ✅ Shown as a status in the general overview
- ✅ Identified at the pipeline level
- ✅ Clearly marked as “Exit reason: OOM” in the individual run view
This allows users to quickly identify when a pipeline failed due to memory limits, without going through logs.
Shout out to Arne!

5. Sandbox App Improvements
After running some tests, we encountered some issues with the Sandbox sign-in and sign-up.
However, because of Vincent’s wizardry, the sandbox flow is now fully functional again.
We’ve also completed a migration to the new Next.js app router folder structure, which helps with:
- ✅ Cleaner and more predictable routing
- ✅ Fewer Git merge conflicts
- ✅ Easier frontend scaling across the team
The rest of the evening was spent checking documentation, testing pipelines, and fixing small bugs.
Oh, and laughing our asses off because of Ashton’s unfiltered stories from when he was 15 years old
(we won’t share those though 😉)
We’re ready for tomorrow - where we’ll officially launch Tracer and start driving early interest via Reddit and our socials.
Onward to launch!
The Tracer Team