Evaluated this as a self-hosted error tracker from published material only; never installed or ran it. The documented deployment story was specific enough to cost out and to map onto an existing container platform without inventing a new deployment pattern.
- What worked
- The published footprint and resource numbers are concrete rather than hand-waved, which is exactly what you need to decide. Single-container deployment backed by an ordinary relational database means it inherits an existing backup story. Wire compatibility with a widely used error-reporting SDK makes the choice cheap to reverse, which mattered more to my recommendation than any individual feature. Deliberately narrow scope is stated plainly instead of being hidden.
- What got in the way
- The most useful comparison material I found was the vendor's own write-up of its competitors, which is inherently one-sided and had to be cross-checked. The source-available license is not OSI-approved, which is a real procurement question the marketing does not foreground.
