Supernode: Interop Acceptance Testing

[axelKingsley]

Note

Because the Interop Activity does not presently do interop, only happy-path testing is implemented here. This sets us up to write the unhappy tests once we're happy with these, while implementing interop algo itself.

Changes

• Introduces a collection of Supernode-Interop focused Acceptance Tests
• Introduces some implementation fixes to support tests:
  • Configuration handling for interopActivationTimestamp works in production launches, but wasn't suitable for devstack. I improved it to handle the config at main, eliminating the issue
  • Interop Activity didn't consider its own activation timestamp when answering if a timestamp is valid. Prior to activation, it must short circuit to "verified", so I did that.
  • Two JSON data marshalling issues were caught by the AI during testing which it fixed. One to use hexutil to accept encoded hex u64s. Another to handle that an internal structure was not a concrete type, and thus needed a special marshalling structure.

Updated Tests

The Supernode Multi-Advance test has been expanded to check Safe Advancement along with Unsafe.

New Tests

TestSupernodeInteropVerifiedAt

Description: Tests that the superroot_atTimestamp endpoint returns correct data after the interop activity has processed timestamps. Uses interop activation at genesis.

Assertions:

• SuperRootAtTimestamp eventually returns non-nil data for the target timestamp (genesis + blockTime)
• No error from SuperRootAtTimestamp when querying the verified timestamp

---

TestSupernodeInteropChainLag

Description: Tests that interop verification is based on SAFE heads, not unsafe heads. Stops chain B's batcher while the CL keeps running, proving the supernode waits for all chains' safe heads before verifying timestamps.

Assertions:

• Both chains advance past block 2 initially
• Baseline timestamp is verified before stopping batcher
• Chain B's safe head stabilizes after batcher is stopped
• Chain B's unsafe head continues to advance even with batcher stopped
• Chain B's safe head remains frozen with batcher stopped
• Timestamps ahead of chain B's frozen safe head are NOT verified (even though chain B's unsafe head is ahead)
• Chain B's safe head catches up after batcher resumes
• Ahead timestamp becomes verified after chain B's safe head catches up
• Both chains continue advancing after recovery

---

TestSupernodeInteropActivationAtGenesis

Description: Tests behavior when interop is activated at genesis time (timestamp 0 offset). Verifies the first timestamps are processed correctly with interop verification from the very beginning.

Assertions:

• SuperRootAtTimestamp eventually returns non-nil data for the first block after genesis (genesis + blockTime)

---

TestSupernodeInteropActivationAfterGenesis

Description: Tests behavior when interop is activated AFTER genesis. Verifies that superroot_atTimestamp returns verified data for timestamps both before and after the activation boundary.

Assertions:

• SuperRootAtTimestamp eventually returns non-nil data for the pre-activation timestamp (genesis + blockTime)
• SuperRootAtTimestamp eventually returns non-nil data for the post-activation timestamp (activation + blockTime)
• Both timestamps verified within 90 seconds

---

TestSupernodeInteropBidirectionalMessages

Description: Tests sending cross-chain messages in both directions (A→B and B→A) to verify the supernode handles bidirectional interop correctly. All messages are valid with no chain interruptions.

Assertions:

• Chain A's safe head advances past the A→B init block
• Chain A's safe head advances past the B→A exec block
• Chain B's safe head advances past the A→B exec block
• Chain B's safe head advances past the B→A init block

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 76.0%. Comparing base (2cfebd1) to head (b1156eb).

Additional details and impacted files

@@            Coverage Diff             @@
##           develop   #18970     +/-   ##
==========================================
- Coverage     76.3%    76.0%   -0.4%     
==========================================
  Files          188      188             
  Lines        10939    10939             
==========================================
- Hits          8357     8324     -33     
- Misses        2436     2471     +35     
+ Partials       146      144      -2     

Flag	Coverage Δ
cannon-go-tests-64	66.4% <ø> (-0.9%)	⬇️
contracts-bedrock-tests	81.7% <ø> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.
see 5 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[axelKingsley]

Because this is mostly generated, I did the first round of review myself. So far I've just reviewed the parts which are not tests to confirm that the wiring isn't doing more than it should. There's also a hack plainly visible in the tests for accessing root RPC (thanks @Inphi for seeing that quickly).

Test review is next, and then I'll consider this complete pending review and move onto interop algo.

[Inphi]

LGTM

[karlfloersch]

The interop message passing test was pretty LEET!