perf: improve performance of the persistent cache used in forking

- Make 'lt' argument more specific in order to retrieve less entries
- Convert block hash to Buffer type immediately on initiatisation (rather than every time we get / put an entry)

src/chains/ethereum/ethereum/src/forking/persistent-cache/persistent-cache.ts

@@ -43,7 +43,7 @@ export class PersistentCache {
  AbstractIterator<Buffer, Buffer>
  >;
  protected ancestry: Ancestry;
- protected hash: Data;
+ protected hashBuffer: Buffer;
  protected request: Request;
  constructor() {}
 
@@ -131,7 +131,7 @@ export class PersistentCache {
  }
 
  async initialize(height: Quantity, hash: Data, request: Request) {
- this.hash = hash;
+ this.hashBuffer = hash.toBuffer();
  this.request = request;
 
  const {
@@ -325,22 +325,21 @@ export class PersistentCache {
  const height = Quantity.from(blockNumber);
  const bufKey = Buffer.from(key);
  const start = lexico.encode([height.toBuffer(), bufKey]);
- const end = lexico.encode([
- Quantity.from(height.toBigInt() + 1n).toBuffer()
- ]);
+ const end = Buffer.concat([start, Buffer.from([0xff])]);
+
  const readStream = this.cacheDb.createReadStream({
  gt: start,
  lt: end,
  keys: true,
  values: true
  });
- const hashBuf = this.hash.toBuffer();
+
  for await (const data of readStream) {
  const { key: k, value } = (data as any) as { key: Buffer; value: Buffer };
  const [_height, _key, blockHash] = lexico.decode(k);
  // if our key no longer matches make sure we don't keep searching
  if (!_key.equals(bufKey)) return;
- if (hashBuf.equals(blockHash) || (await this.ancestry.has(blockHash))) {
+ if (this.hashBuffer.equals(blockHash) || (await this.ancestry.has(blockHash))) {
  return value;
  }
  }
@@ -354,7 +353,7 @@ export class PersistentCache {
  const dbKey = lexico.encode([
  height.toBuffer(),
  Buffer.from(key),
- this.hash.toBuffer()
+ this.hashBuffer
  ]);
  await this.cacheDb.put(dbKey, value);
  return true;