How to efficiently implement auditing for high-throughput background service in .NET 8?

Question

I have a .NET 8 background service responsible for two primary tasks:

1. Push JSON data packets to message brokers (ActiveMQ, RabbitMQ, or JMS queues):

• Sends approximately 25,000 JSON data packets every 2 minutes (data size can vary from 25,000 to 20,000,000).
• This process is handled asynchronously.

2. Manage multiple services using this push mechanism:

• There are 21 different services running concurrently, all working in an asynchronous manner and in single database.

• Each service has a 10-minute sleep interval after sending a packet.

New Requirement:

I now need to implement auditing for every data packet pushed to the broker. Specifically, I need to log each record in a SQL database with a timestamp indicating when it was sent.

Challenges:

• The volume of data being processed is significant, and downtime for the SQL database is not acceptable.
• The solution needs to handle high throughput efficiently without becoming a bottleneck to queue push operations.

Approaches I've considered so far:

1. Implementing a simple fire-and-forget service using C#'s asynchronous capabilities to log the data to SQL. However, I am unsure if this can handle the load effectively.
2. Using a background job processing library like Hangfire to offload the auditing to a separate worker process.

My question is: How can I ensure scalability and reliability while keeping the SQL database from becoming a bottleneck?

I haven't finalized an approach yet and am open to other ideas or improvements. Any suggestions or insights would be greatly appreciated. Thanks!