chore: bump version to 2.0.17

[flyq]

Summary

Bumps the workspace from 2.0.16 to 2.0.17. Every binary still reports 2.0.16 while main carries fifteen merges past the v2.0.16 tag, so the version string cannot be used to tell which code a host is on — the gap that made the fra/ore triage guess at build identity. Cargo.lock regenerated with cargo update -w; no code change. Supersedes #177, which was closed while the frontier-witness P0 work was still outstanding — that work has since landed.

Contents

Everything merged since the v2.0.16 tag (11078d7), oldest first: #167 raw-JSON cache hits · #174 gzip/zstd response compression · #175 direct-from-R2 historical witnesses · #176 concurrent inbound batch entries · #169 malformed tracerConfig rejected in the executor · #172 HTTP connect timeout · #171 single home for block-execution env assembly · #168 resolved (number, hash) threaded through the fetch pipeline · #178 per-request accounting identity and error reasons · #180 deadline give-up names its phase · #182 R2-first witness fetches and a per-hop budget cap · #187 custom-domain R2 target over HTTP/2 · #188 inbound admission control and response-size caps · #190 Rex6 activation in the mainnet genesis · #191 validator R2 witness cap split from the RPC one.

Notes

The v2.0.17 tag is deliberately not part of this PR and still waits on #192. Two items from the above must carry into the release notes and the compatibility table, since both change what an existing deployment has to do:

• fix: add Rex6 activation to mainnet genesis #190 — the validator persists genesis in GENESIS_CONFIG, so an existing DB keeps the pre-Rex6 copy; fixing one means starting once with --genesis-file so it is overwritten. The trace server re-reads the file every start and needs only the replacement.
• refactor: split the validator's R2 witness cap from the RPC one #191 — --r2-max-concurrent-requests is new and now caps R2 GETs. Under --witness-source r2, a host still carrying only the old --witness-max-concurrent-requests spelling is refused at startup by name rather than left uncapped, so env files for r2-mode roles must be migrated before upgrading.

Testing

cargo update -w regenerated the seven workspace-member entries and nothing else; cargo check --workspace clean. No code change.

[mega-maxwell]

Claude review status

Living comment — rewritten in place. The review workflow keeps this single comment up to date instead of posting a new one each round, so it always describes the latest reviewed commit and the earlier text is intentionally gone. No reply is needed here; reply to a finding in its own review thread, and answer an open question in a reply on this PR. The next review round reconciles your answer.

✅ Review clean

Last reviewed: head bfb8d290 · updated 2026-08-31T08:24:26+00:00

New this round: 0 finding(s), 0 question(s) · Resolved this round: 0 · Open questions: 0