Performance and stability of JDeveloper 11gR2 vs. 11gR1

[fjhou...@hotmail.com]

Dear all,

I would like your thoughts on the performance and stability of
JDeveloper 11gR2 vs. 11gR1.

For 5 months now we are building an ADF application in 11gR2
(currently 11.1.2.1) and we daily experience performance and stability
problems with this version of JDeveloper. Although 11gR2 starts up
quickly, compared to the first release we have worked with (11.1.1.3
and 11.1.1.5) we have a strong impression that the 11.1.2.1 release
has proportionally more errors and is less stable. One of our
customers, also developing in 11gR2, experiences the same. Currently,
our project members are discussing and investigating whether it might
be better to go back to the latest 11gR1 version to be certain that
we'll have a stable production environment. Some of our issues
include:

-Many times when we run the project we get 'An unexpected error has
occurred in JDeveloper' error
-Very often during debugging, the debugger goes suddenly to the Timer
class and opens this class in JDeveloper
-Redeploying is still slow, even after we modify the datasource string
as described in: http://andrejusb.blogspot.com/2011/09/how-to-speed-up-application.html
-We have some indications of a memory leak; we see the Java heap size
increasing during each redeployment and often a we get an
outofmemoryerror.
-I've read the comparison between 11gR2 and 11gR1 in
https://forums.oracle.com/forums/thread.jspa?threadID=2329130 , but I
do not read people mentioning the issues we have

Questions:
-Are you experiencing the same performance and stability issues as we
do in 11gR2 ?
-Do you know any ADF applications that are successfully running on a
11gR2 production environment and do perform well for many concurrent
users ?

Regards,
Frank Houweling

[Mark Robinson]

I haven't had any performance issues myself although I have noticed several bugs in JDeveloper.

I, too, frequently get 'An unexpected error has occurred  in JDeveloper' often when I deploy to debug or open the persistence.xml file.

The OutOfMemoryException is a long-time known issue in WLS.  It is particularly noticeable when you use Hudson to do automated builds.

One issue I find really problematic is that JDev caches some of the build data isn't entirely correct in its handling.  Often JDev will not rebuild stale files, this mostly occurs when you do an update from a source control system.

Mark

--
You received this message because you are subscribed to the ADF Enterprise Methodology Group (http://groups.google.com/group/adf-methodology). To unsubscribe send email to adf-methodolo...@googlegroups.com

All content to the ADF EMG lies under the Creative Commons Attribution 3.0 Unported License (http://creativecommons.org/licenses/by/3.0/).  Any content sourced must be attributed back to the ADF EMG with a link to the Google Group (http://groups.google.com/group/adf-methodology).

[shay shmeltzer]

A couple of notes about the out of memory after several deploys.

You might want to try and increase the memory allocated to JDeveloper in the jdev.conf file.

Other parameters that some people found helpful for performance here:

https://blogs.oracle.com/angelo/entry/improving_the_performance_of_jdeveloper 

Also in general with 11.1.2 you shouldn't need to do many redeployments because of the hot-deployment feature - which allows you to save/recompile and reload your page in the browser for most of the changes to existing pages.

Shay

[Renan Monteiro de Castro]

I experienced a lot of problems while trying to run JDeveloper 11.1.2.1 with a 64bit JDK (1.6.31), a good example is that you can't
use the fast swap feature on Weblogic, it will run on a "Cannot attach to a process 32-bits" error.
Also, we experienced less performance and bug issues with the 32-bits version of JDK (both for JDeveloper and Weblogic).

For example, after a SVN Update there is a great chance of some strange compilation error that needs a JDeveloper restart just to delete the .data folder and all its garbage (except for the History files inside it, it saved my life twice).

--
You received this message because you are subscribed to the ADF Enterprise Methodology Group (http://groups.google.com/group/adf-methodology). To unsubscribe send email to adf-methodolo...@googlegroups.com
 
All content to the ADF EMG lies under the Creative Commons Attribution 3.0 Unported License (http://creativecommons.org/licenses/by/3.0/). Any content sourced must be attributed back to the ADF EMG with a link to the Google Group (http://groups.google.com/group/adf-methodology).

--

Renan Monteiro de Castro - Linkedin

[Hyangelo Hao]

The part about WebLogic conking out after several deploys is very true and is one of the many annoying things that sometimes drive me mad about WebLogic and JDeveloper. If I were to be given a dollar every time JDev throws an unexpected exception or weblogic runs out of memory, I'd be able to buy a gizillion more ram to feed my local weblogic server. 

I've gotten used to it by now but when I was starting with the JDev/ADF/WebLogic ecosystem and coming from eclipse/<insert relatively lighter frameworks here>/Tomcat everything felt as slow as molasses. What is worse is that WebLogic takes a while to restart especially if I have some webservices deployed to it. It takes an unholy amount of memory at startup as well. I understand that comparing weblogic with tomcat isn't really fair since Tomcat isn't exactly an application server but Oracle has to do something about this if they want developers to actually like working with ADF. don't get me wrong, I like ADF now. It was an acquired taste for me. But it was a bitter start and the issues outlined above(exceptions, oom, obese development app server) gave me constant, consistent daily grief.

[Oracle ITself]

Hi

i totally agree with you regarding Memory problems in Integrated & Stand Alone Weblogic Servers and Oracle has to do something .. as per my experience in Oracle ADF & Wls .. until now we suffer from this Memory issue ..

Regarding the Stability : 11G R1 still better than R2 until nowm at least from my point of view !!,  some of  problems solved in 11.1.3 & 4 repeated again in 11G R2 !!
and at the same time we face a lot of bugs after we decided to transfer our project from 11.1.1.5 to 11.1.2.0 ..

we contacted Oracle and send them some of theses bugs and they decided that it is really bugs by their Testing Team and many of them will be solved by in coming Release 11.1.2.2 (we are waiting for it now) ..

Note : Oracle will publish new release 11.1.2.2 in coming 2 Months (as per Oracle Team Reply to us) ..

thanks






Wael Abdeen
------------------------------------------------------------------------------------------
Senior Oracle ADF & SOA Consultant at Confidential Company
Kingdom Of Saudi Arabia

Tel          : KSA: +966545874498  Egy : +201150052059
Mail         : Oracle...@Hotmail.com
LinkedIn   : http://sa.linkedin.com/pub/wael-abdeen/12/624/996
FaceBook : http://www.facebook.com/profile.php?id=1272993189

---

Date: Mon, 26 Mar 2012 07:06:43 -0700
From: hyan...@gmail.com
To: adf-met...@googlegroups.com
Subject: [ADF EMG] Re: Performance and stability of JDeveloper 11gR2 vs. 11gR1

[Chris Muir]

Does anyone have bug numbers for the WLS memory problem(s)?

CM.

[Jang Vijay Singh]

Hi,

I have noticed intermittent problems with the debugger which tends to freeze - yet to isolate the issue and reproduce it.

Other than that, jdev 11gR2 perf seems to be good- more so after applying jdec.conf changes suggested earlier in this thread.

There is it seems a known issue with memory leaks on redeployments (which might also impact the integrated instance) but that sorts itself after deleting existing app on wls before redeploying as i found on forums/blogs.

Regards,
JV

Sent from my iPhone

[Hyangelo Hao]

Did my best to search in Oracle Support and the best that I could find is this:   8131207  -  REPETITIVE REDEPLOYING AN EAR USING ANT SCRIPT RESULTS IN OOM

OT: Searching through oracle support reminded me of one other thing that Oracle definitely needs to do to improve ADF: expose the bug parade publicly even if it is just read only. Make it indexable and searchable through google. That should help developers find out easier if the issue they are seeing is a known issue or not since the first reasonable place to start searching would be on google. Searching for a bug in Oracle Support is quite unsavory and a pain in the ass. I wonder if Oracle bothered applying their 'usability expertise' on that service? 

[Chris Muir]

Thanks for searching for the bug. Note the bug is specific to Ant
deployments to WLS. It also looks like the bug was originally
recorded in BEA's systems and when ported to Oracle bugs database, it
was graded as severity 3. It hasn't been looked at again since then.

Unless a more recent or relevant bug is found and posted here, my
recommendation to the community, collaborate to log a new reproducible
bug which is specific to JDev IDE deployments WLS, share the bug
number on this group and then separately log SRs against that bug
stating you want it fixed. With the weight of a new bug, several
customers asking for fixes etc, I'll *try* and action something
internally. This issue has been talked about for so long it would be
nice to get it off the radar.

As for the quality of Oracle Support services, this is beyond the
scope of this group ;-)

CM.

[Jan Vervecken]

hi

- about "Does anyone have bug numbers for the WLS memory problem(s)?"
-- Today I noticed bug 13964205, "WLS GETS OOME IF ADF APPLICATION IS
DEPLOYED REPEATEDLY."

regards
Jan Vervecken

On Mar 27, 1:58 am, Chris Muir <chriscm...@gmail.com> wrote:
> Does anyone have bug numbers for the WLS memory problem(s)?
>
> CM.
>

> On 26 March 2012 22:21, Oracle ITself <oracle_its...@hotmail.com> wrote:
>
>
>
>
>
>
>
>
>
> > Hi
>
> > i totally agree with you regarding Memory problems in Integrated & Stand
> > Alone Weblogic Servers and Oracle has to do something .. as per my
> > experience in Oracle ADF & Wls .. until now we suffer from this Memory issue
> > ..
>
> > Regarding the Stability : 11G R1 still better than R2 until nowm at least
> > from my point of view !!,  some of  problems solved in 11.1.3 & 4 repeated
> > again in 11G R2 !!
> > and at the same time we face a lot of bugs after we decided to transfer our
> > project from 11.1.1.5 to 11.1.2.0 ..
>
> > we contacted Oracle and send them some of theses bugs and they decided that
> > it is really bugs by their Testing Team and many of them will be solved by
> > in coming Release 11.1.2.2 (we are waiting for it now) ..
>
> > Note : Oracle will publish new release 11.1.2.2 in coming 2 Months (as per
> > Oracle Team Reply to us) ..
>
> > thanks
>
> > Wael Abdeen
> > ------------------------------------------------------------------------------------------
> > Senior Oracle ADF & SOA Consultant at Confidential Company
> > Kingdom Of Saudi Arabia
>
> > Tel          : KSA: +966545874498  Egy : +201150052059

> > Mail         : Oracle_ITs...@Hotmail.com

> > LinkedIn   :http://sa.linkedin.com/pub/wael-abdeen/12/624/996
> > FaceBook :http://www.facebook.com/profile.php?id=1272993189
>
> > ________________________________
> > Date: Mon, 26 Mar 2012 07:06:43 -0700

> > From: hyang...@gmail.com

> >https://forums.oracle.com/forums/thread.jspa?threadID=2329130, but I

[Chris Muir]

Thanks for the follow up Jan.

This bug links to the none-published bug 13726772 which is flagged as
no longer reproducible in the upcoming 11.1.2.2.0. The analyst still
needs to confirm bug 13964205 is fixed against 11.1.2.2.0 and I'll let
them run with it for a few days to give them time to do their work and
not double up effort on my part.

I know you'll ask why bug 13726772 was not published. In this case
the bug was logged against a future internal JDev build and Oracle's
policy is not to publish internal build bugs, just those on public
releases.

I've recorded this bug as issue ADFEMG-16 in Jira such that we can
follow it's progress from here: http://java.net/jira/browse/ADFEMG-16

CM.

[Hyangelo]

This is also happening in 11.1.1.4 fyi. Without exaggeration, this is one of my top productivity sink holes on a day to day basis. It isn't like WLS restarts quickly, with a couple WebServices deployed to it, my WLS takes 5-7 minutes to restart. In a single work day, I estimate that I do around 4-5 WLS restarts which translates to losing half an hour. At first this really drove me nuts but over time I learned to adjust and just find (productive) other things to do while WLS is restarting.

P.S. Is it really necessary to some make ADF/JDev/WLS bugs private/unpublished? I think it this particular practice is unfair to us developers and making the bug parade public and searchable from google will be a really big help.

[Chris Muir]

Understood, in fact I even blogged the same sometime back about Google. But suggesting such in this forum won't get change to Oracle's support policies.

CM.

On 17/04/2012, at 9:16 PM, Hyangelo <hyan...@gmail.com> wrote:

P.S. Is it really necessary to some make ADF/JDev/WLS bugs private/unpublished? I think it this particular practice is unfair to us developers and making the bug parade public and searchable from google will be a really big help.

--

[Jan Vervecken]

hi Hyangelo

- about "... this is one of my top productivity sink holes on a day to
day basis ..."
-- If you think you are hitting bug 13964205 you could express your
interest to have it fixed by leaving a comment at http://java.net/jira/browse/ADFEMG-16
or in an Oracle service request.

- about "... making the bug parade public and searchable ..."
-- Although new, currently http://java.net/projects/adfemg and its
JIRA issue tracker is the closest to that there is I think.

regards
Jan Vervecken

On Apr 17, 3:16 pm, Hyangelo <hyang...@gmail.com> wrote:
> > This is also happening in 11.1.1.4 fyi. Without exaggeration, this is one
> > of my top productivity sink holes on a day to day basis. It isn't like WLS
> > restarts quickly, with a couple WebServices deployed to it, my WLS takes
> > 5-7 minutes to restart. In a single work day, I estimate that I do around
> > 4-5 WLS restarts which translates to losing half an hour. At first this
> > really drove me nuts but over time I learned to adjust and just find
> > (productive) other things to do while WLS is restarting.
>

> P.S. Is it really necessary to *some* make ADF/JDev/WLS bugs

[Hyangelo]

Yes, I agree. I don't think posting here or in OTN or anywhere can change Oracle's policy. Personally, this is one of the primary downside of using ADF as opposed to other open-source alternatives. 

[prasad narahari]

All
We have an ADF application developed and running on 11.1.1.2, Our external customers are reporting IE9 browser compatability issues occasionally. So, we have decided to upgrade to either 11.1.1.6 or 11.1.2.2.
After reading the release notes of 11.1.2.2, I was more inclined towards 11.1.2.2 as Oracle is saying there are performance gains due to javascript optimizations etc.,

My search for 11.1.2.2 in this Google group resulted in this thread which is talking about performance issues and stability issues.
Many people mentioned these problems being part of 11.1.2.1. I am curious to hear if any body has any experiences with 11.1.2.2

Not sure if it is worth going for 11.1.2.2 against 11.1.1.6 at this point of time

Also within 11gR1 which would you prefer 11.1.16 or 11.1.15

One more thing, our application (11.1.1.2) is using web services (WSDLs) out of a seperate 11.1.1.5 SOA suite. They are running on different web logic instances.

Appreciate your feedback, this would really help us.

--

You received this message because you are subscribed to the ADF Enterprise Methodology Group (http://groups.google.com/group/adf-methodology). To unsubscribe send email to adf-methodolo...@googlegroups.com
 
All content to the ADF EMG lies under the Creative Commons Attribution 3.0 Unported License (http://creativecommons.org/licenses/by/3.0/). Any content sourced must be attributed back to the ADF EMG with a link to the Google Group (http://groups.google.com/group/adf-methodology).

--
Thanks
Prasad
Cell:510-565-5008

[Vik]

Well my advise would be to to 11.1.2.2 if you are planning to migrate. There may be little more hiccups but it would be worth for performance gains and more features delivered in 11.1.2.2

Thankx and Regards

Vik
Founder
http://www.sakshum.org
http://blog.sakshum.org

[Chad Thompson]

On Monday, July 30, 2012 at 9:01 PM, prasad narahari wrote:

Not sure if it is worth going for 11.1.2.2 against 11.1.1.6 at this point of time

This will depend on your immediate needs:  11.1.2.2 does have a few enhancements and is the 'bridge' to the next major releases of ADF and Fusion technologies that will use the JEE6 standards (Facelets for view layers, etc.)

With the 11.1.1.x versions you should be able to simply 'upgrade' your JDeveloper (and the ADF libraries on your deployment environments!) without too much trouble - the 11.1.2.x versions are a little different in terms of patching, use of Enterprise Manager/Fusion Middleware Control, etc.

Also within 11gR1 which would you prefer 11.1.16 or 11.1.15

I'd jump to the 'highest version you can support' because the newer versions will have bug fixes, minor enhancements, etc. that will make things easier (hopefully) for you.  Which brings me to the last point:

One more thing, our application (11.1.1.2) is using web services (WSDLs) out of a seperate 11.1.1.5 SOA suite. They are running on different web logic instances.

The other thing to consider here is your developer/development environment - if you are supporting both the SOA Suite plus the ADF application, you might find it easier to match ADF and SOA versions for your deployment and development, if anything to keep "one version of JDeveloper" and minimize any issues with incompatibility.

If most of your development is Fusion Middleware focused, you might have an easier time sticking with the 11.1.1.x versions until you do a larger migration to the 11.1.2.x / 12c versions of ADF and Fusion Middleware.

- Chad

-- 
Chad Thompson 
chad_t...@mac.com

http://chadthompson.me