incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gary Dusbabek <gdusba...@gmail.com>
Subject Re: Multi-indexing data
Date Mon, 29 Mar 2010 12:46:44 GMT
It sounds like you might need a main storage CF and several CFs to
serve as inverted indices to support querying.  The inverted indices
basically map the searchable attribute (as a key) to the row id
(column name) of the main storage.  Keep in mind that the searchable
attribute may need to map to multiple ids in the main storage.

Since the ability to query by attribute seems to be a goal of yours,
you might want to look into Lucandra, which is a melding of the Lucene
indexer with Cassandra storage.

Gary.


On Mon, Mar 29, 2010 at 01:45, Matthew Stump <mrevilgnome@gmail.com> wrote:
> Hi, I have a question about Cassandra's data model I was hoping you guys could help me
with.  Most of our queries are performed against a series of tables containing crypto keys
and their associated meta data.  A key could have any number of identifiable attributes that
need to be searchable: iasn, 64bit key id, 32bit key id, expiration, revoker etc…  From
what I understand I don't believe tagging the same information with multiple keys is supported.
 The best that I could think of was to add key/value pairs in the form "keyid_32_0x73A5DC55:some
junk data" to the row or maintain a separate set of columns that provide "keyid_32_0x73A5DC55:key
row name" mappings and perform two queries.  I'm not a fan of either of these options.  Is
there some other solution that I may have overlooked?
>
> Thanks,
> -Matt

Mime
View raw message