HBase: row key size

I come from the background of RDBMS and recently started reading HBase. I understand that there are no secondary indexes, and we should not try to do something like:

SELECT * FROM tbl_photo WHERE album_id = 1969

I was wondering if all the information can be used to create the string itself.

For example, There is a user with his / her electronic registers with a photo-sharing service provider. The user can create an album (create several albums) and upload photos to it. Another user comments on photos, and some users vote for the comment.

The key to determining such a vote may look like email:album:ts:photo:ts:comment:ts:vote:ts. Does this key make sense? Longer than recommended? ( ts stands timestamp)

+5
source share
1 answer

, , ? ? HBase - , HBaseCon , : HBase Ian Varley.

, , HBase: " ?".

( ) , - email:album, , ,...

, , , . , , email:picture , -. email:album:picture, , ().

, , , 2000 , email:picture email:album album picture , HBase 2000 , .

Depending on which version of HBase you are using, you may implement some kind of secondary index using FuzzyRowFilter .

+7
source

All Articles