[
https://issues.apache.org/jira/browse/AVRO-853?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13060970#comment-13060970
]
Doug Cutting commented on AVRO-853:
-----------------------------------
It's okay to use schemas as Map keys, but it's an error to change a schema
after putting it in a Map. We don't detect that error. If schema's were
immutable that error would be impossible.
I'm not too worried about using zero to indicate 'invalid hash'. A more random
value might be a bit better, but a collision with this value is not fatal, it
just makes the cache ineffective for those few schemas whose hashcode is that
value. We could instead have a flag to indicate that the hashcode is unset, or
use a boxed Integer, both of which would use more memory without benefit in
2^32-1 cases. Whatever value we use should probably be a constant. We can
explicitly check when calculateHashCode returns this value and use a different
value when it occurs.
In summary, I think adding a hashcode cache to Schema's would be a good thing
to do. I think this patch could be improved a bit, but I'm generally in favor
of committing something like this soon.
> Cache hash codes in Schema and Field
> ------------------------------------
>
> Key: AVRO-853
> URL: https://issues.apache.org/jira/browse/AVRO-853
> Project: Avro
> Issue Type: Improvement
> Components: java
> Affects Versions: 1.5.1
> Reporter: Douglas Kaminsky
> Attachments: AVRO-853.patch
>
>
> We are experiencing a serious performance degradation when trying to
> store/retrieve fields and schemas in hash-based data structures (eg.
> HashMap). Since all fields and schemas are immutable (with the exception of
> RecordSchema allowing deferred setting of Fields) it makes sense to cache the
> hash code on the object instead of recalculating every time the hashCode
> method gets called.
> (Are there other mutable Schema sub-types that I'm not thinking about?)
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
