That is incorrect. You can have the used space filewalker disabled and still get the benefit from the badger cache.
After adding
storage2.monitor.dedicated-disk: true
storage2.monitor.reserved-bytes: 100 GB
storage2.piece-scan-on-startup: false
to all nodes, which got updated to v1.114.6 the multinode dashboard shows 0B used. The other columns look ok. MND is on v1.114.6 too.
I got 1 Windows node upgraded automatically to 114.6 and now lazy filewalker entry on config.yaml is gone.
Is that to be expected?
Kind of expected. The used space will be always off until all places would be modified to support this hidden experimental feature.
It’s not possible as far as I know, the config file is generated only on install. It doesn’t modify the existing one.
This is easy to check - add it back, downgrade to something like 1.114.5 and start only the storagenode-updater
service, the storagenode
service should be stopped, wait a little bit, and it will update the storagenode binary to the latest correct version and start it, then you can check the config file after that.