[
https://issues.apache.org/jira/browse/AVRO-1836?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15266822#comment-15266822
]
Ryan Blue commented on AVRO-1836:
---------------------------------
[[email protected]], it looks like you probably used the same Schema parser
for both strings, which would end up attempting to redefine the schema because
the parser keeps track of Schemas it has parsed so that others can reference
them by name. I don't think your error is coming from the compatibility check.
If you have a stack trace, that would show where the exception is coming from.
> avro schema compatibility checking does not work for records.
> -------------------------------------------------------------
>
> Key: AVRO-1836
> URL: https://issues.apache.org/jira/browse/AVRO-1836
> Project: Avro
> Issue Type: Bug
> Affects Versions: 1.8.0
> Reporter: Yan Cui
> Priority: Blocker
>
> Hi Guys,
> We found a problem on avro 1.8.0 when calling avro's compatibility checking.
> Specifically, we used the function checkReaderWriterCompatibility,
> it works with no problem when we provides simple schema, like
> oldschema={"type":"int"} schema={"type":"long"}, but when we use complicated
> schema, like
> oldschema={"type":"record","name":"twitter","fields":[{"name":"username","type":"string"}]}
> schema={"type":"record","name":"twitter","fields":[{"name":"username","type":"string"}]}
> it reports that record name(twitter in this case) cannot be redefined.
> When one rename the second schema, it told me the two schema are not
> compatible. Actually, for the completed case, oldschema and schema are
> exactly the same, the output is expected to be compatible.
> Do you guys have any ideas about this?
> Thanks, Yan
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
