Commit 5b328222 authored by Dan Notestein's avatar Dan Notestein
Browse files

Add docs for utilities (beginning with compress_block_log).

parent 778e24fe
# Documentation for auxiliary programs useful for hived nodes
## compress_block_log: compresses/decompresses block_log files
### Example usage for compress_block_log
Create compressed version of blocklog in blockchain dir in blockchain2 dir:
`compress_block_log ./datadir/blockchain ./datadir/blockchain2`
Same as above, with explicitly specified arguments (non-positional):
`compress_block_log -i ./datadir/blockchain -o ./datadir/blockchain2`
Created compressed blocklog with 5000 blocks using 8 threads:
`compress_block_log -j8 -n5000 ./datadir/blockchain ./datadir/blockchain2`
Decompress blocklog in blockchain dir to blockchain2 dir:
`compress_block_log --decompress ./datadir/blockchain ./datadir/blockchain2`
### Allowed options for compress_block_log
--decompress Instead of compressing the block log,
decompress it
--zstd-level arg (=15) The zstd compression level to use
(-131072 - 22)
--benchmark-decompression decompress each block and report the
decompression times at the end
-j [ --jobs ] arg (=1) The number of threads to use for
-i [ --input-block-log ] arg The directory containing the input
block log (required)
-o [ --output-block-log ] arg The directory to contain the compressed
block log (required)
--dump-raw-blocks arg A directory in which to dump raw,
uncompressed blocks (one block per
-s [ --starting-block-number ] arg (=1)
Start at the given block number (for
benchmarking only, values > 1 will
generate an unusable block log)
-n [ --block-count ] arg Stop after this many blocks
-h [ --help ] Print usage instructions
### Overview of compress_block_log
The compress_block_log tool requires a valid block_log.index file.
If no block_log.index file is present, it will build one first before
beginning the compression/decompression step.
If the block_log.index is valid but too long, the block_log.index will be
truncated to the number of blocks in the blocklog. If the block_log.index is
too short, the block_log.index will be extended to match the size of the blocklog.
When using positional arguments, explicitly specified options must come first.
When compressing a block_log, it is recommended to use -j set to the number
of cpus available to avoid long compression times.
This tool also replaces the previous `truncate_block_log` utility. To truncate
a blocklog, see the third example above using the -n option.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment