Is there any document talking about Defender operation architecture? I want to know how Defender, esp Sentinel, make sure that it never miss contract event detections

Is there any document talking about Defender operation architecture? I want to know how Defender, esp Sentinel, makes sure that it never misses contract event detections

1 Like

Hi @masakore a high level explanation of our Sentinel architecture is as follows:
We are essentially continuously polling for new blocks on every Defender supported network. When we see a new block we process all the transactions in the block. Once the transactions are processed we match them to sentinels according to well defined transaction conditions. Once matched, we have additional logic that determines if an attached autotask should run / should a notification be sent.
Internally, we have extensive monitoring and well defined procedures in place to quickly identify and resolve any critical failures that occur within our pipeline.

I hope that helps :slight_smile:

2 Likes