[
https://issues.apache.org/jira/browse/UIMA-941?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12749523#action_12749523
]
Adam Lally commented on UIMA-941:
---------------------------------
In AnalysisEngineImplBase.buildProcessTraceFromMBeanStats, it is specifically
checking for a time > 0 before adding an event. We could remove that check.
I think the intention of that check might have been for remote services where
"analysis time" is not available (but "service call time" is), and so it was
better to not log an analysis time than to log an artificial time of 0. A way
to fix that would be to set times originally to -1 [indicating not applicable]
and change them to 0 the first time any time is logged. A simpler alternative
might be in the case where both service call time and analysis time are 0, to
create an Analysis entry in the ProcessTrace, with a time of 0.
> No ProcessTrace events for process calls that take 0ms
> ------------------------------------------------------
>
> Key: UIMA-941
> URL: https://issues.apache.org/jira/browse/UIMA-941
> Project: UIMA
> Issue Type: Bug
> Components: Core Java Framework
> Affects Versions: 2.2.2
> Environment: Windows XP, Java 1.5
> Reporter: Thilo Goetz
> Priority: Minor
>
> Found when testing release candidate 4 of 2.2.2. Here's the text of an email
> I wrote. Adam responded that it's probably a bug.
> http://www.mail-archive.com/uima-dev%40incubator.apache.org/msg06565.html
> I was just testing CVD for the next release, and tried
> out the new performance report feature. Internally, this
> uses the ProcessTrace feature of AnalysisEngine.process().
> It seems that if the process() call returns very quickly,
> i.e., in 0ms according to the Java timer, no ProcessTrace
> sub-events are recorded. For example, I'm running the
> tagger, and here's some output I get:
> This one looks fine, as expected:
> 3/27/08 4:29:52 PM - 10:
> org.apache.uima.tools.cvd.MainFrame.internalRunAE(1570): INFO: Process trace
> of AE run:
> Component Name: HmmTaggerTAE
> Event Type: Analysis
> Duration: 2078ms (100%)
> Sub-events:
> Component Name: WhitespaceTokenizer
> Event Type: Analysis
> Duration: 265ms (12.75%)
> Component Name: Hidden Markov Model - Part of Speech Tagger
> Event Type: Analysis
> Duration: 1813ms (87.25%)
> Component Name: Fixed Flow Controller
> Event Type: Analysis
> Duration: 0ms (0%)
> Next one seems to have lost the tagger:
> 3/27/08 4:30:08 PM - 10:
> org.apache.uima.tools.cvd.MainFrame.internalRunAE(1570): INFO: Process trace
> of AE run:
> Component Name: HmmTaggerTAE
> Event Type: Analysis
> Duration: 16ms (100%)
> Sub-events:
> Component Name: WhitespaceTokenizer
> Event Type: Analysis
> Duration: 16ms (100%)
> Component Name: Fixed Flow Controller
> Event Type: Analysis
> Duration: 0ms (0%)
> Finally, this one only has the flow controller:
> 3/27/08 4:30:19 PM - 10:
> org.apache.uima.tools.cvd.MainFrame.internalRunAE(1570): INFO: Process trace
> of AE run:
> Component Name: HmmTaggerTAE
> Event Type: Analysis
> Duration: 0ms
> Sub-events:
> Component Name: Fixed Flow Controller
> Event Type: Analysis
> Duration: 0ms
> Looking at this in the debugger, there are no subevents
> there, so it's not a problem with the printing.
> Is this expected, or is it a bug? Seems to me it is a bug,
> but I'm not sure from the documentation.
> --Thilo
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
