The batch fossilizer currently stores fossils locally (using files) before periodically sending a bitcoin transaction.
This doesn't work well in a micro-services architecture because a deployment or auto-scaling can screw up the files and make us lose pending-fossilization data (or worse fossilized evidence that hasn't reached the store).
We need to decouple the batch fossilizer from its storage mechanism. This way we can use an implementation that leverages AWS SQS to make sure we never lose data.
The batch fossilizer currently stores fossils locally (using files) before periodically sending a bitcoin transaction. This doesn't work well in a micro-services architecture because a deployment or auto-scaling can screw up the files and make us lose pending-fossilization data (or worse fossilized evidence that hasn't reached the store). We need to decouple the batch fossilizer from its storage mechanism. This way we can use an implementation that leverages AWS SQS to make sure we never lose data.