How modest are we talking?
You can start with just 1 CPU core, 1 GB of RAM, and 40 GB of disk. For a busier setup, 2 cores and 4 GB of RAM give Telebugs more room to breathe. That is the modest server we used for the published benchmark.
A modest server is plenty
Telebugs is measured on ordinary VPS hardware and designed to do focused error tracking without demanding an infrastructure team.
admin@telebugs:~/telebugs$ bin/load --users 100 --duration 10m --read-ratio 20 \
--groups 200 --drain-timeout 3600
Published 10-minute benchmark
---------------------------------
ServerHetzner CX23
Hardware2 vCPU / 4 GB RAM / 40 GB local disk
Workload80% error ingestion / 20% authenticated UI reads
Accepted53.15/s
Processed end-to-end50.08/s
HTTP p952.11sIs 50 reports a second just a burst number?
No. We kept the test running for 10 minutes, used the UI at the same time, and waited for Telebugs to finish everything still in the queue. The 50.08 figure counts completed reports, not requests that merely reached the front door. You can see the whole benchmark.
Will every workload fit?
No, and they do not have to. Error volume, retention, release artifacts, and how many projects share the server all add weight. Start small. If Telebugs needs more room later, move it to a bigger server. Your apps will not know the difference.
What about ARM?
It works there too. Telebugs supports AMD64 and ARM64. An ARM cloud server or Raspberry Pi-class machine is fine, as long as it has enough RAM and disk for the amount of error data you keep.