Adrian Sutton
Adrian Sutton
@totocz A 2 second arrival is relatively slow - ideally it should be under 1s, but it's not generally the end of the world. The main problem with those numbers...
The late arrivals are almost always because the block producer (someone else's validator) was slow in producing the block. There's nothing you can do about that.
I'm going to close this one as we've ironed out the initial performance issues with Teku a while back and what's left just boils down to besu import times which...
> I'm wondering if it's not worth to have timestamp as a mandatory field in all of the reorgs we've had from mainnet having the timestamp was instrumental to quickly...
Yeah, I acknowledge it's useful, but I wouldn't want to require fork choice tracks it just for this API, hence I think it should be optional. Clients can then choose...
Can confirm that we're consistently seeing full blocks now. Important note that EIP1559 targets an average of 50% full blocks. There can be variation around that so it's not unexpected...
Ah actually the: ``` {"@timestamp":"2022-10-16T10:05:11,856","level":"INFO","thread":"nioEventLoopGroup-3-3","class":"ForwardSyncStep","message":"Saved blocks 7776535 -> 7776577 (target: 7778836)","throwable":""} ``` lines towards the end are more useful when you can spot them in all the backwards sync noise....
So to provide some context... Execution clients use devp2p for connections between nodes and discovery v4 for the UDP based node discovery. Both of those only support IPs and DNS...
> Teku for example has a similar bug right now, and others have had them in the past. This wasn't for lack of a spec, and the particular request doesn't...
> JSON is a data interchange format above all - it happens to be readable which is helpful sometimes, but users would not be looking at this data (which user...