fix: store blockhash alongside blocks

[alexghr]

Small PR to store a block's hash in the archiver. This should speed things up when doing small operations like getting a transaction from a block since logs don't have to deserialized and added to a block.

Related #3870

[AztecBot]

Benchmark results

Metrics with a significant change:

• batch_insert_into_append_only_tree_ms (2): 16.1 (-88%)
• batch_insert_into_indexed_tree_ms (512): 359 (-31%)
• node_database_size_in_bytes (10): 3,728,927 (-30%)

Detailed results

All benchmarks are run on txs on the Benchmarking contract on the repository. Each tx consists of a batch call to create_note and increment_balance, which guarantees that each tx has a private call, a nested private call, a public call, and a nested public call, as well as an emitted private note, an unencrypted log, and public storage read and write.

This benchmark source data is available in JSON format on S3 here.

Values are compared against data from master at commit 5fe99085 and shown if the difference exceeds 1%.

L2 block published to L1

Each column represents the number of txs on an L2 block published to L1.

Metric	8 txs	32 txs	128 txs
l1_rollup_calldata_size_in_bytes	45,444	179,588	716,132
l1_rollup_calldata_gas	221,064	866,444	3,447,548
l1_rollup_execution_gas	934,208	3,968,918	24,791,084
l2_block_processing_time_in_ms	1,360 (-10%)	5,116	20,955 (-1%)
note_successful_decrypting_time_in_ms	305 (-3%)	932 (-3%)	3,358 (-3%)
note_trial_decrypting_time_in_ms	14.1 (-36%)	59.1 (-36%)	138 (-4%)
l2_block_building_time_in_ms	15,123 (+2%)	59,335	237,595
l2_block_rollup_simulation_time_in_ms	10,754 (+2%)	42,037	168,682
l2_block_public_tx_process_time_in_ms	4,336	17,200	68,645

L2 chain processing

Each column represents the number of blocks on the L2 chain where each block has 16 txs.

Metric	5 blocks	10 blocks
node_history_sync_time_in_ms	14,866 (-4%)	28,999 (-2%)
note_history_successful_decrypting_time_in_ms	2,165 (-4%)	4,576 (+3%)
note_history_trial_decrypting_time_in_ms	105 (-30%)	179 (-6%)
node_database_size_in_bytes	3,622,521	⚠️ 3,728,927 (-30%)
pxe_database_size_in_bytes	29,923	59,478

Circuits stats

Stats on running time and I/O sizes collected for every circuit run across all benchmarks.

Circuit	circuit_simulation_time_in_ms	circuit_input_size_in_bytes	circuit_output_size_in_bytes
private-kernel-init	213	44,681	25,561
private-kernel-ordering	165	43,753	14,809
base-rollup	1,067	128,338	881
root-rollup	80.7 (-3%)	4,088	709
private-kernel-inner	291	71,208	25,561
public-kernel-private-input	205	31,863	25,561
public-kernel-non-first-iteration	202	31,905	25,561
merge-rollup	8.37 (-1%)	2,608	881

Tree insertion stats

The duration to insert a fixed batch of leaves into each tree type.

Metric	1 leaves	2 leaves	8 leaves	16 leaves	32 leaves	64 leaves	128 leaves	512 leaves	1024 leaves	2048 leaves	8192 leaves
batch_insert_into_append_only_tree_ms	11.1 (-1%)	⚠️ 16.1 (-88%)	13.6 (+2%)	17.9 (-1%)	23.8 (+1%)	49.7 (-1%)	77.5 (-15%)	243	472	912	3,567
batch_insert_into_indexed_tree_ms	N/A	N/A	N/A	64.4 (-1%)	N/A	59.2 (-1%)	130 (+17%)	⚠️ 359 (-31%)	734 (-1%)	1,408	5,550

Miscellaneous

Transaction sizes based on how many contracts are deployed in the tx.

Metric	0 deployed contracts	1 deployed contracts
tx_size_in_bytes	15,455	32,224

Transaction processing duration by data writes.

Metric	0 new commitments	1 new commitments
tx_pxe_processing_time_ms	371 (-1%)	1,080

Metric	1 public data writes
tx_sequencer_processing_time_ms	534

[spalladino]

I'm worried that this change could mean that some consumers of L2Blocks will no longer be receiving their blocks. WDYT about having L2Block and L2BlockWithLogs as two separate interfaces, so it's absolutely clear when we are moving around an object with logs, and when it doesn't?

[just-mitch]

Not having an interface for L2Block is strange to me since it seems a major point of interaction for clients. Having two separate interfaces for blocks with/without logs feels like it would get painful as well. Maybe just:

interface L2Block {
    ...
    newEncryptedLogs?: L2BlockL2Logs,
    ...
}
type L2BlockWithLogs = L2Block & { newEncryptedLogs: L2BlockL2Logs };
type L2BlockWithoutLogs = L2Block & { newEncryptedLogs: undefined };

then in the class implementation of L2Block.

hasLogs(): this is L2BlockWithLogs {
    return this.newEncryptedLogs !== undefined
}

[alexghr]

I agree with both of you, but changes are coming to this section of the code to bring it inline with the Yello paper and I'd prefer to wait for them to land before touching it more than needed if that's alright :)

[spalladino]

Judging from this change, wouldn't it be easier to just squeeze blockHash as another of the fields of L2Block? Or does it break serialization to/from noir?

[just-mitch]

It looks like this is planned as part of #3938 ?

aztec-packages/yarn-project/circuit-types/src/l2_block.ts

Line 358 in f81b7c0

// TODO(#3938): populate bodyHash

[alexghr]

I can't find it now but there was an issue/PR about blocks now being serialisable because they didn't logs.

I'd consider doing this but only if I make blockHash mandatory in the constructor so it doesn't cause any errors downstream 😬