cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gary Dusbabek (JIRA)" <>
Subject [jira] Commented: (CASSANDRA-1814) validation is inefficient
Date Tue, 07 Dec 2010 23:05:02 GMT


Gary Dusbabek commented on CASSANDRA-1814:

Some of the validations (integer, bytes) just become noops.  Most just needed to verify field
length.  Utf8 was tricky.  Instead of using a CharsetDecoder (which would have generated garbage),
I looked for a validation algorithm that operated on the raw bytes.

> validation is inefficient
> -------------------------
>                 Key: CASSANDRA-1814
>                 URL:
>             Project: Cassandra
>          Issue Type: Improvement
>    Affects Versions: 0.7.0 rc 1
>            Reporter: Gary Dusbabek
>            Assignee: Gary Dusbabek
>         Attachments: v1-0001-type-validations-that-generate-less-garbage.txt
> We do all validation by simply calling AbstractType.getString().  This generates garbage
needlessly and has a lot of overhead.
> A simpler interface would be to make AbstractType.validate abstract and have the child
classes implement it in an intelligent and efficient way.
> EDIT: Somewhat related:  It looks like we're attempting to validate column names in ThriftValidation.validateColumns().
 Is this intentional?  Nevermind that part. I get it.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message