mirror of
https://github.com/XRPLF/rippled.git
synced 2025-12-06 17:27:55 +00:00
Enhance nudb_block_size documentation with detailed performance characteristics
- Expand 4096 bytes description with memory footprint and latency details - Add sequential throughput and I/O operation details for 8192-16384 bytes - Include enterprise environment recommendations for 32768 bytes - Provide comprehensive guidance for block size selection based on hardware
This commit is contained in:
@@ -994,17 +994,22 @@
|
||||
# traditional filesystems (ext4, NTFS, HFS+).
|
||||
# Provides good balance of performance and storage
|
||||
# efficiency. Recommended for most deployments.
|
||||
# Minimizes memory footprint and provides consistent
|
||||
# low-latency access patterns across diverse hardware.
|
||||
#
|
||||
# - 8192-16384 bytes: May improve performance on
|
||||
# high-end NVMe SSDs and copy-on-write filesystems
|
||||
# like ZFS or Btrfs that benefit from larger block
|
||||
# alignment. Can reduce metadata overhead for large
|
||||
# databases.
|
||||
# databases. Offers better sequential throughput and
|
||||
# reduced I/O operations at the cost of higher memory
|
||||
# usage per operation.
|
||||
#
|
||||
# - 32768 bytes (32K): Maximum supported block size
|
||||
# for high-performance scenarios with very fast
|
||||
# storage. May increase memory usage and reduce
|
||||
# efficiency for smaller databases.
|
||||
# efficiency for smaller databases. Best suited for
|
||||
# enterprise environments with abundant RAM.
|
||||
#
|
||||
# Note: This setting cannot be changed after database
|
||||
# creation without rebuilding the entire database.
|
||||
|
||||
Reference in New Issue
Block a user