ETL can't proceed when writing big Batch

[godexsoft]

In Clio we have Batched writes and the size of the batch is not limited by anything. If it happens that there are a big amount of transactions in one ledger we end up hitting the threshold of Scylla for maximum batch data size in Kb.
When this happens, ETL can't progress.

Example of a ledger that got this issue: https://devnet.xrpl.org/ledgers/3058257

Suggested fix is to split writes into smaller batches which is to be controlled by a configuration parameter.