diff options
author | Viktor Söderqvist <viktor.soderqvist@est.tech> | 2022-11-09 18:50:07 +0100 |
---|---|---|
committer | GitHub <noreply@github.com> | 2022-11-09 19:50:07 +0200 |
commit | 4e472a1a7fc0dc2c7da2b48ac7342e9385b4f92a (patch) | |
tree | 62715e6a8fb51264449aa6ef0866f92d9d1e00bf /src/aof.c | |
parent | 07d187066a86a9a3124af740373bf5bf49e345ff (diff) | |
download | redis-4e472a1a7fc0dc2c7da2b48ac7342e9385b4f92a.tar.gz |
Listpack encoding for sets (#11290)
Small sets with not only integer elements are listpack encoded, by default
up to 128 elements, max 64 bytes per element, new config `set-max-listpack-entries`
and `set-max-listpack-value`. This saves memory for small sets compared to using a hashtable.
Sets with only integers, even very small sets, are still intset encoded (up to 1G
limit, etc.). Larger sets are hashtable encoded.
This PR increments the RDB version, and has an effect on OBJECT ENCODING
Possible conversions when elements are added:
intset -> listpack
listpack -> hashtable
intset -> hashtable
Note: No conversion happens when elements are deleted. If all elements are
deleted and then added again, the set is deleted and recreated, thus implicitly
converted to a smaller encoding.
Diffstat (limited to 'src/aof.c')
-rw-r--r-- | src/aof.c | 67 |
1 files changed, 21 insertions, 46 deletions
@@ -1818,56 +1818,31 @@ int rewriteListObject(rio *r, robj *key, robj *o) { * The function returns 0 on error, 1 on success. */ int rewriteSetObject(rio *r, robj *key, robj *o) { long long count = 0, items = setTypeSize(o); - - if (o->encoding == OBJ_ENCODING_INTSET) { - int ii = 0; - int64_t llval; - - while(intsetGet(o->ptr,ii++,&llval)) { - if (count == 0) { - int cmd_items = (items > AOF_REWRITE_ITEMS_PER_CMD) ? - AOF_REWRITE_ITEMS_PER_CMD : items; - - if (!rioWriteBulkCount(r,'*',2+cmd_items) || - !rioWriteBulkString(r,"SADD",4) || - !rioWriteBulkObject(r,key)) - { - return 0; - } + setTypeIterator *si = setTypeInitIterator(o); + char *str; + size_t len; + int64_t llval; + while (setTypeNext(si, &str, &len, &llval) != -1) { + if (count == 0) { + int cmd_items = (items > AOF_REWRITE_ITEMS_PER_CMD) ? + AOF_REWRITE_ITEMS_PER_CMD : items; + if (!rioWriteBulkCount(r,'*',2+cmd_items) || + !rioWriteBulkString(r,"SADD",4) || + !rioWriteBulkObject(r,key)) + { + return 0; } - if (!rioWriteBulkLongLong(r,llval)) return 0; - if (++count == AOF_REWRITE_ITEMS_PER_CMD) count = 0; - items--; } - } else if (o->encoding == OBJ_ENCODING_HT) { - dictIterator *di = dictGetIterator(o->ptr); - dictEntry *de; - - while((de = dictNext(di)) != NULL) { - sds ele = dictGetKey(de); - if (count == 0) { - int cmd_items = (items > AOF_REWRITE_ITEMS_PER_CMD) ? - AOF_REWRITE_ITEMS_PER_CMD : items; - - if (!rioWriteBulkCount(r,'*',2+cmd_items) || - !rioWriteBulkString(r,"SADD",4) || - !rioWriteBulkObject(r,key)) - { - dictReleaseIterator(di); - return 0; - } - } - if (!rioWriteBulkString(r,ele,sdslen(ele))) { - dictReleaseIterator(di); - return 0; - } - if (++count == AOF_REWRITE_ITEMS_PER_CMD) count = 0; - items--; + size_t written = str ? + rioWriteBulkString(r, str, len) : rioWriteBulkLongLong(r, llval); + if (!written) { + setTypeReleaseIterator(si); + return 0; } - dictReleaseIterator(di); - } else { - serverPanic("Unknown set encoding"); + if (++count == AOF_REWRITE_ITEMS_PER_CMD) count = 0; + items--; } + setTypeReleaseIterator(si); return 1; } |