feat(aggregation mode): add retry logic to batches download

[JuArce]

feat(aggregation mode): add retry logic to batches download

Description

This PR adds retry logic to the download of batches on the aggregation mode

The retry parameters are

// Retry parameters for S3 requests/// Initial delay before first retry attempt (in milliseconds)constRETRY_MIN_DELAY_MILLIS:u64 = 500;/// Exponential backoff multiplier for retry delaysconstRETRY_FACTOR:f32 = 2.0;/// Maximum number of retry attemptsconstRETRY_MAX_TIMES:usize = 5;/// Maximum delay between retry attempts (in seconds)constRETRY_MAX_DELAY_SECONDS:u64 = 10;

How to Test

1. Run Ethereum Package

make ethereum_package_start

2. Run batcher

make batcher_start_ethereum_package

3. Send tasks

make batcher_send_proof_with_random_address

4. Run aggregation mode

make proof_aggregator_start_ethereum_package AGGREGATOR=sp1

In this case, the aggregation mode should be able to fetch the batch from Localstack.

5. Stop Localstack

6. Run aggregation mode

make proof_aggregator_start_ethereum_package AGGREGATOR=sp1

In this case, you should see the process tries to fetch the batch, waiting before doing a retry using exponential backoff,

2025-08-05T18:22:09.684662Z INFO proof_aggregator::backend::s3: Fetching batch from S3 URL: http://localhost:4566/aligned.storage/628809ec4e40fd964766a0fe33e47d6779816b411a8310e1c6202fe00f2d7675.json
2025-08-05T18:22:09.685407Z WARN proof_aggregator::backend::s3: Failed to send request to http://localhost:4566/aligned.storage/628809ec4e40fd964766a0fe33e47d6779816b411a8310e1c6202fe00f2d7675.json: error sending request for url (http://localhost:4566/aligned.storage/628809ec4e40fd964766a0fe33e47d6779816b411a8310e1c6202fe00f2d7675.json)
2025-08-05T18:22:10.187615Z INFO proof_aggregator::backend::s3: Fetching batch from S3 URL: http://localhost:4566/aligned.storage/628809ec4e40fd964766a0fe33e47d6779816b411a8310e1c6202fe00f2d7675.json
2025-08-05T18:22:10.189572Z WARN proof_aggregator::backend::s3: Failed to send request to http://localhost:4566/aligned.storage/628809ec4e40fd964766a0fe33e47d6779816b411a8310e1c6202fe00f2d7675.json: error sending request for url (http://localhost:4566/aligned.storage/628809ec4e40fd964766a0fe33e47d6779816b411a8310e1c6202fe00f2d7675.json)

7. Start Localstack again. In this case, the storage will be available, but the batch will not exist. This is a permanent error.

8. Run aggregation mode

make proof_aggregator_start_ethereum_package AGGREGATOR=sp1

In this case, the error is permanent, so the process will not retry. You should see the following logs:

2025-08-05T18:27:32.397180Z ERROR proof_aggregator::backend::fetcher: Error while downloading proofs from s3. Err StatusFailed((404, "Not Found"))

Type of change

• New feature

Checklist

• “Hotfix” to testnet, everything else to staging
• Linked to Github Issue
• This change depends on code or research by an external entity
  • Acknowledgements were updated to give credit
• Unit tests added
• This change requires new documentation.
  • Documentation has been added/updated.
• This change is an Optimization
  • Benchmarks added/run
• Has a known issue
  • Link to the open issue addressing it
• If your PR changes the Operator compatibility (Ex: Upgrade prover versions)
  • This PR adds compatibility for operator for both versions and do not change crates/docs/examples
  • This PR updates batcher and docs/examples to the newer version. This requires the operator are already updated to be compatible