init.prune=full, the latest 128 blocks are too few.

[d11myx]

Hi, as the ARB full node runs, the disk space is getting bigger and bigger. I want to enable pruning --init.prune=full, but I found that the document states that only the latest 128 blocks will be retained. Is this too little? Is there a parameter to adjust the number of latest blocks to be retained?

If I understand it wrongly, please correct me, thank you.

[joshuacolvin0]

Note that pruning only happens offline, it does not prune while listening for RPC requests. This means that the node will need to be stopped regularly to prune the database.
https://docs.arbitrum.io/run-arbitrum-node/run-full-node#pruning

Pruning a full node only keeps the last snapshot. Historical queries are handled by data built up in memory cache, so archive data isn't stored in database. You can control how much data is kept in memory cache using --execution.caching.block-count