[
https://issues.apache.org/jira/browse/UIMA-1040?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Marshall Schor updated UIMA-1040:
---------------------------------
Affects Version/s: 2.2.2AS
Fix Version/s: (was: 2.2.2AS)
not fixing this for 2.2.2AS release
> Modify dd2spring to handle UIMA-AS Aggregate Cas Multipliers
> ------------------------------------------------------------
>
> Key: UIMA-1040
> URL: https://issues.apache.org/jira/browse/UIMA-1040
> Project: UIMA
> Issue Type: Bug
> Components: Async Scaleout
> Affects Versions: 2.2.2AS
> Reporter: Jerry Cwiklik
> Priority: Minor
>
> dd2spring determines if a component is a Cas Multiplier by checking the
> component's descriptor outputsNewCASes property. If the component is a CM,
> the dd2spring generates spring xml artifacts to create a special cas pool to
> be used to create new CASes. For a Primitive CM, this is necessary. As the
> Primitive CM actually creates new CASes from an input CAS it needs the
> special cas pool. If the CM is an Aggregate, this special cas pool is not
> necessary. The aggregate CM *outputs* new CASes, but it doesnt create them.
> It uses a shadow CAS Pool (one per remote CM delegate) to deserialize CASes
> instead. The current dd2spring code generates a warning when parsing a
> deployment descriptor containing an aggregate CM:
> *** WARN: line-number: 34 deployment descriptor for analysisEngine:
> <key>Queue is for a CAS Multiplier (or Collection Reader wrapped as a CAS
> Multiplier). However, the <casMultiplier> element is missing.
> Defaulting to a poolSize of 1 for this case.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
