Genesis sync stalls at L1 batch 16529 — need guidance or a DB dump

[littlehand26]

1. What happens

Using docker/.env.mainnet's full-archive config (EN_SNAPSHOTS_RECOVERY_ENABLED=false, EN_PRUNING_ENABLED=false) on official matterlabs/external-node:v29.14.0, the external node fails at L1 batch 16529 with this error
(byte-for-byte reproducible across runs):

Task state_keeper failed: cannot reject transaction 0x76866def29a0...
Account validation error: function_selector = 0x03eb8b54
required: 50,835,401,000,000 wei
provided: 47,129,162,082,376 wei

Earlier in the same batch, tx 0xeae48a77... (which calls EVM emulator contract 0x00005ea0... introduced in V27) replays as success/gasUsed=451558 on our EN vs. revert/gasUsed=248544 on mainnet. V26→V27 upgrade landed at batch
16375.

We also tried eth.drpc.org archive RPC as L1 client (contradicting the suggestion in #20) — same crash, byte-identical wei values.

2. Ask

We need full-archive history (txs / traces / events from genesis) for our downstream pipeline. Snapshot recovery works but discards pre-snapshot history.

Could you either:

1. Provide a pg_restore DB dump of an archive node (like matter-labs' en-backups.matterlabs.dev for zkSync Era), or
2. Document the correct way to perform a genesis sync — we may be missing a configuration step or workaround.