Peter Broadhurst

That’s certainly the behaviour as designed today. That those inputs a base 10. The new FFTM connector toolkit is more permissive, using the utility helper FFBigInt from firefly-common. A PR...

Yep - I think it uses the “json.Number” string type today

Hi @anujmax , EthConnect is the raw connector for Ethereum chains (see #149 for where the community is up to on chains with less predictable behaviors than BFT chains with...

Hi @feiyiban, you can certainly use Docker to run EthConnect (and all of Hyperledger FireFly) on Windows. However, for compilation and development a Linux VM will be needed currently. A...

This is affect v0.12.0 of FireFly, which shipped this change: https://github.com/hyperledger/firefly/commit/c07090310dffe024aa5ca193e51533df0988e027

It seems to be the case that as well as resetting, the event stream does not receive events after the initial batch that are found during the rewind.

Ok - found the root cause here. It's subtle: - The `catchupModePageSize` was set to be GREATER THAN the `catchupModeBlockGap` - This meant our logic at the point we caught...

Note I found we did still have a problem with writing incremental HWM checkpoint while doing gap based recovery: https://github.com/hyperledger/firefly-ethconnect/pull/199#issuecomment-1029662339

So there's a timing here when the `Message already in-flight` error message does not come out, and the duplicate actually gets processed - because the in-flight is flushed before the...

The problem was we were relying on async timing to validate duplicate detection. So in https://github.com/kaleido-io/ethconnect/pull/65 I've added a separate synchronous test for duplicate detection in the Kafka bridge, and...