Does Couchbase store JSON as plain text in disk? -
i know json returned not same stored (properties names may not came in same order, example).
but, how couchbase stores json objects in disk? takes advantage of schema or writes simple plain text?
in case of huge json's, there drawback of compressing , store binary? (besides loosing query capacity, , so).
in addition wiki page mr llama linked to, great , not going rehash hear, here other answers questions.
if pre-compress documents in app, of course move on wire faster , through queues faster if compressed. downside 1 pointed out, lose query ability. no indexes, no views, no n1ql. binary object. if have great key pattern application using , going going after each object key , never query go it! couchbase great.
all being said, couchbase snappy compression files on disk. saves on disk space, not on wire or in flight.
just aware couchbase working on new persistence engine called forestdb replace current couchstore. https://github.com/couchbase/forestdb
on note, if have huge jsons, might worthwhile normalizing them bit. has make sense app, if can, pulling parts of information need, when need them , can optimize use of couchbase's managed cache. example, in user profile store might need user's authentication information of time , in own tiny document , want in managed cache time, user's security questions need , if not in managed cache , couchbase has fetch them disk, big deal. if in large user profile document, have fetch entire object including possibly lot of data not need right moment. proper standardized key patterns, can data need or use bulk , of documents.
Comments
Post a Comment