Picon
Favicon

[jira] Commented: (UIMA-1130) Deployment Descriptor should allow setting the number of concurrent listeners for a reply queue


    [
https://issues.apache.org/jira/browse/UIMA-1130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619006#action_12619006
] 

Marshall Schor commented on UIMA-1130:
--------------------------------------

Good points, Adam, to consider.

Here are a couple of others (from reading our docs): are there use cases where the remote delegate would be
able to get requests from the remote broker, but possibly not be able to send it replies?  (e.g., due to
firewall issues?)

Would it be possible to set the prefetch value for reply queues to be some fixed number (e.g., 1 or 10 or ... ?)
There should only ever be one listener pulling messages off of the reply queue.  Is there a penalty for
setting up the prefetch value to something high?

> Deployment Descriptor should allow setting the number of concurrent listeners for a reply queue
> -----------------------------------------------------------------------------------------------
>
>                 Key: UIMA-1130
>                 URL: https://issues.apache.org/jira/browse/UIMA-1130
>             Project: UIMA
>          Issue Type: Improvement
>          Components: Async Scaleout
>    Affects Versions: 2.2.2
>            Reporter: Adam Lally
>
> The Spring XML allows setting a concurrentConsumers property for a reply queue (either an aggregate's
(Continue reading)

Picon
Favicon

[jira] Commented: (UIMA-1130) Deployment Descriptor should allow setting the number of concurrent listeners for a reply queue


    [
https://issues.apache.org/jira/browse/UIMA-1130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619008#action_12619008
] 

Eddie Epstein commented on UIMA-1130:
-------------------------------------

The capability to add listeners for a remote reply queue should give equal or better performance than
setting a prefetch value in most cases. Can we see if a single tuning parameter is enough before adding more
complexity? Note that prefetch is not part of the JMS standard and is not available in all JMS implementations.

>are there use cases where the remote delegate would be able to get requests from the remote broker, but
possibly not be able to send it replies? (e.g., due to firewall issues?) 
Yes, when the client is behind a firewall, but the remote delegate [service] and it's broker are outside the
firewall. This is one of the motivations for always allocating the reply queue on the service's broker
(the other main one being to eliminate the need for a colocated broker to instantiate a local reply queue,
again something not possible with many JMS implementations).

>There should only ever be one listener pulling messages off of the reply queue. 
It is sometimes desirable to have more than one thread doing deserialization of reply messages, which is
the original point of this issue.

>Is there a penalty for setting up the prefetch value to something high?
The main problem for UIMA is memory management, as serialized XmiCas messages can be quite large.

> Deployment Descriptor should allow setting the number of concurrent listeners for a reply queue
> -----------------------------------------------------------------------------------------------
>
>                 Key: UIMA-1130
(Continue reading)

Jörn Kottmann | 1 Aug 20:33
Picon

Re: Cas Viewer discussion

> Good suggestion to have a new plugin to switch the underline tools  
> (CAS
> Editor or CAS Viewer). We just provide some customization (through  
> Eclipse
> preference ?) for users to select what is the "default" mode they  
> would like
> to have since different users may have different usage pattern.
>
> Actually, we did take this "integrated" approach to have only ONE  
> tool to
> edit UIMA component descriptors and the "deployment descriptor" used  
> in
> UIMA-AS deployment. That is the CDE which is used to edit all UIMA XML
> descriptors.

Actually for the Cas Viewer case we need two plugins.

1. One plugin which contains editors which can show and manipulate
a CAS file. This would be for now the CAS Editor and the CAS Viewer,
later also the CAS journaling.

2. In your case a new plugin which takes care of searching the type  
system
in the users project folder and providing style information. The type  
system
and style information are then passed to the first plugin to open the  
actual
editor.  The users chooses which editor is preferred with the eclipse  
default editor
setting for CAS files. I think we can call this uima dev plugin.
(Continue reading)

Adam Lally | 1 Aug 20:40
Picon
Gravatar

Re: Cas Viewer discussion

On Fri, Aug 1, 2008 at 2:33 PM, Jörn Kottmann <kottmann@...> wrote:
> I still think that the huge benefit of doing it the Cas Editor way with
> multiple views:
>
> + Its easier to explore a CAS file because there is more screen space to see
> things.
>

I agree with that.  I've used the CAS Editor a few times, and
basically always need to maximize that editor window in order to see
anything.  It basically always needs to take over all of the screen
real-estate in order to be useful to me, so it might as well be its
own perspective.
  -Adam

Picon
Favicon

[jira] Commented: (UIMA-1130) Deployment Descriptor should allow setting the number of concurrent listeners for a reply queue


    [
https://issues.apache.org/jira/browse/UIMA-1130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619143#action_12619143
] 

Marshall Schor commented on UIMA-1130:
--------------------------------------

Thanks, Eddie.  Let me clarify my statements a bit :-)...

>>>The capability to add listeners for a remote reply queue should give equal or better performance than
setting a prefetch value in most cases. Can we see if a single tuning parameter is enough before adding more
complexity? Note that prefetch is not part of the JMS standard and is not available in all JMS
implementations. 

+1 to avoiding more tweaking/tuning parameters whenever possible, in favor of something that just always
works well.

>>>> are there use cases where the remote delegate would be able to get requests from the remote broker, but
possibly not be able to send it replies? (e.g., due to firewall issues?)
>>>Yes, when the client is behind a firewall, but the remote delegate [service] and it's broker are outside
the firewall. This is one of the motivations for always allocating the reply queue on the service's broker
(the other main one being to eliminate the need for a colocated broker to instantiate a local reply queue,
again something not possible with many JMS implementations).

Well, I meant the other way 'round.  So let me try again - are there use cases where the remote broker serves the
remote delegate new work OK, but for some reason can't host the reply queue?

>>>>>There should only ever be one listener pulling messages off of the reply queue.
>>>It is sometimes desirable to have more than one thread doing deserialization of reply messages, which
(Continue reading)

Picon
Favicon

[jira] Created: (UIMA-1134) Put additional scaleout scenarios in UIMA AS getting started

Put additional scaleout scenarios in UIMA AS getting started
------------------------------------------------------------

                 Key: UIMA-1134
                 URL: https://issues.apache.org/jira/browse/UIMA-1134
             Project: UIMA
          Issue Type: Improvement
          Components: Async Scaleout
            Reporter: Eddie Epstein
            Assignee: Eddie Epstein

--

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Picon
Favicon

[jira] Closed: (UIMA-1134) Put additional scaleout scenarios in UIMA AS getting started


     [
https://issues.apache.org/jira/browse/UIMA-1134?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Eddie Epstein closed UIMA-1134.
-------------------------------

    Resolution: Fixed

Added scenarios as suggested by Greg Holmberg

> Put additional scaleout scenarios in UIMA AS getting started
> ------------------------------------------------------------
>
>                 Key: UIMA-1134
>                 URL: https://issues.apache.org/jira/browse/UIMA-1134
>             Project: UIMA
>          Issue Type: Improvement
>          Components: Async Scaleout
>            Reporter: Eddie Epstein
>            Assignee: Eddie Epstein
>

--

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Burn Lewis (JIRA | 4 Aug 16:58
Picon
Favicon

[jira] Updated: (UIMA-1131) Delegates following a CAS Multiplier do not get the complete typesystem


     [
https://issues.apache.org/jira/browse/UIMA-1131?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Burn Lewis updated UIMA-1131:
-----------------------------

    Attachment: UIMA-1131.temp.patch

Problem caused by too-early creation of the CasMultiplier's CasPool, before the all the typesystems have
been merged.  This patch removes some of the changes made in uimaj-as-core-UIMA-1104-patch-10.txt so
that the pool is again created only when the first Cas is requested.  Presumably this will cause a problem
when the JMX monitor tries to monitor the CM's non-existent CasPool, but that will have to be fixed another way.

> Delegates following a CAS Multiplier do not get the complete typesystem
> -----------------------------------------------------------------------
>
>                 Key: UIMA-1131
>                 URL: https://issues.apache.org/jira/browse/UIMA-1131
>             Project: UIMA
>          Issue Type: Bug
>          Components: Async Scaleout
>            Reporter: Burn Lewis
>         Attachments: UIMA-1131-testcase.patch, UIMA-1131.temp.patch
>
>
> Delegates after a CAS multiplier get only the type system of the first delegate declared in the aggregate. 
CASes generated by a CAS multiplier don't get the full merged type system.

--

-- 
(Continue reading)

Picon
Favicon

[jira] Commented: (UIMA-1131) Delegates following a CAS Multiplier do not get the complete typesystem


    [
https://issues.apache.org/jira/browse/UIMA-1131?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619627#action_12619627
] 

Marshall Schor commented on UIMA-1131:
--------------------------------------

Please create a new Jira for the JMX issue so we don't forget to figure out a fix for that.

> Delegates following a CAS Multiplier do not get the complete typesystem
> -----------------------------------------------------------------------
>
>                 Key: UIMA-1131
>                 URL: https://issues.apache.org/jira/browse/UIMA-1131
>             Project: UIMA
>          Issue Type: Bug
>          Components: Async Scaleout
>            Reporter: Burn Lewis
>         Attachments: UIMA-1131-testcase.patch, UIMA-1131.temp.patch
>
>
> Delegates after a CAS multiplier get only the type system of the first delegate declared in the aggregate. 
CASes generated by a CAS multiplier don't get the full merged type system.

--

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

(Continue reading)

Marshall Schor | 4 Aug 23:35

http://issues.apache.org/jira/browse/UIMA-1002 ?

It may be just me (it's late), but this issue 
http://issues.apache.org/jira/browse/UIMA-1002 was created, then closed 
with a fix indicated in the comment, but there is no svn update or 
patch?  Maybe the subversion commits didn't get posted (this is supposed 
to happen automatically, but sometimes the mechanism fails).

I don't know how to search thru all of the svn history easily, looking 
for any mention of UIMA-1002... 

Does anyone remember any details about this particular fix?

-Marshall


Gmane