Sequence Numbers - SpiderRock/SRDataFeed GitHub Wiki

The SpiderRock feed contains a sequence number in the message header. This is a byte, and just cycles through 255 and starts over again all day long. The purpose of it is to catch any network issues where a packet may have been lost (due to the non-error correcting aspect of multicast).

The sequence number should be tracked via this key: multicast channel, source_id, message_type

So the same multicast channel and source_id have different sequence numbers based on message_type. This is more granular then the exchanges, which only have sequence numbers based on a multicast channel. The SpiderRock sequence numbers allows you to track whether, for instance, a print message was lost vs a quote message, etc.