Fjord Upgrade of opBNB¶

Hardfork

Upgrade Timeline¶

The Fjord upgrade will happen at:

Testnet: Sep-10-2024 06:00 AM +UTC
Mainnet: Sep-24-2024 06:00 AM +UTC

Upgrade opBNB op-node and op-geth to v0.5.0 Before Hardfork¶

Releases:

Docker Images

ghcr.io/bnb-chain/op-node:v0.5.0
ghcr.io/bnb-chain/op-geth:v0.5.0
ghcr.io/bnb-chain/op-batcher:v0.5.0
ghcr.io/bnb-chain/op-proposer:v0.5.0

Key Highlight: L1 fee calculation changed¶

Fjord updates the L1 cost calculation function to use a FastLZ-based compression estimator. The L1 cost is computed as:

l1FeeScaled = l1BaseFeeScalar*l1BaseFee*16 + l1BlobFeeScalar*l1BlobBaseFee
estimatedSizeScaled = max(minTransactionSize * 1e6, intercept + fastlzCoef*fastlzSize)
l1Fee = estimatedSizeScaled * l1FeeScaled / 1e12

You can find the detailed explanation in the spec.

User Facing Changes¶

New flag --wait-node-sync added to op-batcher (default false), indicates if during startup, the batcher should wait for a recent batcher tx on L1 to finalize (via more block confirmations). This should help avoid duplicate batcher txs
New flag --wait-node-sync added to op-proposer (default false), indicates if during startup, the proposer should wait for the rollup node to sync to the current L1 tip before proceeding with its driver loop
New flag --compression-algo added to op-batcher (default zlib), user can choose brotli algo after Fjord fork
New flag --l1.rpc-max-cache-size added to op-node (default 1000), so user can config the the maximum cache size of the L1 client

Pull Requests¶

Merge upstream v1.7.7 by @bnoieh in https://github.com/bnb-chain/opbnb/pull/216
feat(op-node): Keep consistent status when meet an unexpected el sync by @krish-nr in https://github.com/bnb-chain/opbnb/pull/222
feat(op-node): add l1 cache size config by @welkin22 in https://github.com/bnb-chain/opbnb/pull/225
feat(op-chain-ops): add Wright fork config into genesis file generation code by @welkin22 in https://github.com/bnb-chain/opbnb/pull/226