uima-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Spico Florin <spicoflo...@gmail.com>
Subject Some questions about CASPool functionality
Date Wed, 08 Feb 2012 08:30:26 GMT
Hello!
   I would like to know how CASPool works for an AE, regarding the
following aspects:
 - when a CAS is returned to the CASPool of the AE? After is being
processed by *ALL* Analysis Engines in the pipeline or after is being
consumed by the CASConsumer?
For example:
I have a pipeline that is running parallel AEs in pipeline and one
CASConsumer (using the AdvancedFlowController)
<nameValuePair>
        <name>Flow</name>
        <value>
          <array>
              <string>OAnnotator,LAnnotator,RAnnotator
             </string>
            <string>CCASConsumer</string>
          </array>
        </value>
      </nameValuePair>

Each of those annotators has its own CASPool with different sizes. Also
OAnnotator_process_time < RAnnotator_process_ time <
LAnnotator_process_time.
 Suppose that we have one message processed by the above pipeline. In this
use case how the CASPool will behave for each of the AE? It will take into
the consideration the whole time processing
(*ALL* the AE finished to send the annotations to the consumer) or we have
the same order kept for putting CAS into the CASPool :
OAnnotator_CAS_put_in_Pool_time < RAnnotator_CAS_put_in_Pool_time <
CAS_put_in_Pool_time.

 I look forward for your answers.
  Thank you.


Regards,
  Florin

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message