Hello Collection, Consensus and Verification node operators,
Currently, the collection, consensus and verification nodes intermittently experience spikes in disk usage where the node software consumes 10-15% more disk space than average for a few minutes before returning back to the average usage. The spikes in disk usage are caused by the compaction process of the Badger database used by the node software.
Please monitor the node disk usage and ensure it has room to accommodate those spikes. If needed, please add more disk space for the node without deleting the existing data. The Flow Foundation nodes are currently running with the following disk sizes and are being upgraded if disk usage goes beyond 80%.
Collection node: 400 GB
Consensus node: 150 GB
Verification node: 150 GB
We expect these spikes will be eliminated once the Badger database is replaced by the Pebble database in the future, but until then, we request you proactively monitor your node disk usage.
Please let us know if you have any questions.
Thank you,
Flow Foundation