greg/pcompress
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

More tweaks to slightly improve segment dedupe efficiency.

Browse source 
Use on average 8MB segments for all cases.
Some minor cleanps.
This commit is contained in:
Moinak Ghosh 2013-04-24 19:13:07 +05:30
parent eabd670790
commit 5d6ffd969d
3 changed files with 4 additions and 17 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
rabin/global/dedupe_config.c
View file

@ -185,8 +185,8 @@ read_config(char *configfile, archive_config_t *cfg)
cfg->verify_chunks = 0;
cfg->algo = COMPRESS_LZ4;
cfg->chunk_cksum_type = DEFAULT_CHUNK_CKSUM;
cfg->similarity_cksum = DEFAULT_SIMILARITY_CKSUM;
cfg->pct_interval = DEFAULT_SIMILARITY_INTERVAL;
cfg->similarity_cksum = GLOBAL_SIM_CKSUM;
cfg->pct_interval = DEFAULT_PCT_INTERVAL;
fh = fopen(configfile, "r");
if (fh == NULL) {
@ -272,13 +272,6 @@ read_config(char *configfile, archive_config_t *cfg)
fclose(fh);
return (1);
}
} else if (strncmp(line, "SIMILARITY_CKSUM", 16) == 0) {
cfg->chunk_cksum_type = get_cksum_type(pos);
if (cfg->chunk_cksum_type == CKSUM_INVALID) {
fprintf(stderr, "Invalid CHUNK_CKSUM setting.\n");
fclose(fh);
return (1);
}
}
}
fclose(fh);
@ -373,10 +366,6 @@ set_config_s(archive_config_t *cfg, const char *algo, cksum_t ck, cksum_t ck_sim
cfg->dedupe_mode = MODE_SIMPLE;
cfg->segment_sz_bytes = user_chunk_sz;
cfg->similarity_cksum_sz = cfg->chunk_cksum_sz;
} else if (cfg->archive_sz < (ONE_TB * 100)) {
cfg->segment_sz_bytes = FOUR_MB;
} else {
cfg->segment_sz_bytes = EIGHT_MB;
}

4
rabin/global/dedupe_config.h
View file

@ -33,11 +33,9 @@
extern "C" {
#endif
#define DEFAULT_SIMILARITY_INTERVAL 5
#define DEFAULT_CHUNK_CKSUM CKSUM_SHA256
#define DEFAULT_SIMILARITY_CKSUM CKSUM_BLAKE256
#define DEFAULT_COMPRESS COMPRESS_LZ4
#define DEFAULT_PCT_INTERVAL 5
#define DEFAULT_PCT_INTERVAL 4
#define CONTAINER_ITEMS 2048
#define MIN_CK 1
#define MAX_CK 5

2
rabin/global/index.c
View file

@ -167,7 +167,7 @@ set_cfg:
*pct_interval = 0;
} else {
cfg->intervals = 100 / *pct_interval;
cfg->sub_intervals = (cfg->segment_sz / cfg->intervals) >> 1;
cfg->sub_intervals = cfg->intervals;
*hash_slots = file_sz / cfg->segment_sz_bytes + 1;
*hash_slots *= cfg->sub_intervals;
}