CPerezz Articles

Twenty-one months of mainnet data on why the first slot of every epoch is reorged ~7× more than the rest — and where Glamsterdam should set the attestation deadline.

Identical EEST bloatnet runs report wildly different MGas/s on three geth databases. The cause turned out to be a 380 MiB pathdb journal shipped inside one snapshot — a provenance artifact, not a property of either database.

The same experiment on Besu. Two of the three causes are artifacts of how the store was built — where the pre-run's rows land in the LSM tree, and a generated store written with no bloom filters at all.

The same experiment on Nethermind, and the first one taken all the way. The 17× is where the rows sit, not what they contain: equalise placement and re-run the full suite, and every category that reads state comes back to parity.