I'm creating a web page to monitor bitcoin mempool https://mempoolexplorer.com/ and I've realized that sometimes exists huge transactions dependency graphs in the mempool.
I'm talking about +-700 transactions which depends among them, and although no mempool rule is broken, i.e. max mempool ancestors for a transaction, I'm puzzled about what can be doing this kind of dependency graphs.
I've stored some graphs in json format: https://gist.github.com/dev7ba/c144c68127b97082652bc860cc95edf6 and https://gist.github.com/dev7ba/50167d6e336100698ab2599123444efc that contains some examples of those graphs.
Those gists .json are created by the backend of my web page that you can call it here: https://mempoolexplorer.com/txmempool/miningQueueAPI/tx/{txIdHere} to obtain a transaction statistics currently in mempool. See txDependenciesInfo.nodes
and txDependenciesInfo.edges
data to obtain the dependency info.
Also, as it's difficult to obtain a transaction with dependencies, I've created a section https://mempoolexplorer.com/txsGraphs or https://mempoolexplorer.com/txmempool/miningQueueAPI/txGraphList which lists all the dependency graphs that exists currently in the mempool, with the transactions set, and whether the graph is linear or not (i.e. using Child Pays For Parent several times instead of BIP-125).
Be aware that mempool is refreshed in the back-end each 5 seconds. But if the mempool is too large: >30000 transactions, graphs calculation will be omitted to avoid high overload.
Thanks.