Oc4j not fully utilizing all cpus

We are running some performance testing in anticipation of deploying our application to production.
We are doing quite a lot of XSL transformations in our app -- the user logs in, we use their id to find the proper XML record (about 8k in size) stored in our database, then using the XML record we retrieve two XSLTs also stored in the database (80kb and 24kb). We have a static util XSLHelper class which receives the XML plus the first XSLT as input (Strings) which returns an XML string, then we call the XSLHelper again with that XML plus the other XSLT as input. Finally the resulting string output is HTML which we display to the user.
During our testing, we noticed that one CPU is pegged to 100% while the other is barely used at all. It seems like Oc4j is only spinning the XSLHelper calls off into a single cpu rather than splitting the load across both cpus.
I'm curious if anyone has any experience with this kind of thing and can share how they resolved the issue, so both CPUs were used as desired?
Thanks.

Just set the number of processes (default islands) for your OC4J to 2. The setting can be found at the server properties of your container in the enterprise manager website.

Similar Messages

Fully utilizing all CPU's for a LabVIEW application

Hi All,
Have any of you figured out how to harness all of the CPU's in modern machines?
Backgroud:
I have an application that does a lot of signal processing and it was pegging the CPU of the machine it was originally deployed on for many minutes.
As a quick first step we suggested the customer try the application on a new high-end machine. THey did and the preformance improved ...
BUT...
When we look at the
Task Manager >>> Perfomance tab
it appears we are not not utilizing all of the available CPU's.
This observation is based on the 8 CPU graphs displayed in the Task Manager.
The first 4 graphs show very heavy CPU useage but the reamining four graphs show little or no loads.
I am guessing that this may be due to LV (8.X) using a default of 4 threads for each execution system.
Since the last time we were on-site, I have looked at
...\LabVIEW\vi.lib\utility\sysinfo.llb\threadconfig.vi
and it appears all I have to do is run that utility one time and save the config as 8 threads for each execution system.
Now before I send someone back to site, I'd like to find out if someone has traveled this road before me and would like to share their wisdom.
Thank you,
Ben
"Mommy, I want to go FAST!" (Daughter of one of my old girl friends)
Ben Rayner
I am currently active on.. MainStream Preppers
Rayner's Ridge is under construction
Solved!
Go to Solution.

I received a couple of questions concerning this post, so a bit more information to clear things up.
LabVIEW's Logic
LabVIEW's default thread creation logic is to create the max of (number of cores, 4). The bug we had is that we inadvertently limited this to four (no humorous comments please ).
What does the utility do?
The utility writes some settings to your ini file if different than default. When you run the utility, it displays the number of processors you have in a field at the top. It displays the number of threads being used for each priority in a section below. For a machine with 1-4 cores, the default thread count will be four. For a machine with 8 cores, the default thread count "should" be 8 (and, as of LabVIEW 8.5, it is).
Roy

WLS slow while machine not fully utilized

Hello,
can anybody please help us with pointers on where to look in case of
the following symptoms:
We have WLS 5.1SP8 on HPUX, jvm 1.2.2.04 connecting to Oracle and
Informix
The WLS seems to be slow, even when the CPU utilisation is around 50%.
It never goes to
full 100%. It very much seems like a configuration issue, the question
just is which one...
The tiers in front and behind also report low cpu utilisation.
Current config:
max 80 concurrent requests allowed by client.
Oracle database pool 128 (I know it is too large, planning to reduce)
maxthreads 96
Weblogic oracle OCI driver
many thanks in advance
madis

While I do not agree that, in general, 96 threads is too many (it may or may
not be and this is very dependent on the hardware/OS/JVM/application -- I will
say that I have never used more than about 125 even with applications running
very high transaction rates), I do think that Charlie is on the right track.
A quick test that I always do to help isolate the problem is to drop one of our
example servlets (e.g., SessionServlet) into the server and run the load test
against this. Since it does no backend access and very little CPU work, it
lets you see if your are able to push requests into the server fast enough to
fully utilize the server. Often times, this is where the problem lies and is
much easier to isolate than application/database issues (unless you've built
your code in such a way that you can measure and/or short-circuit pieces of
functionality).
Once you verify that you are able to push enough requests into the server to
fully utilize it, then I would go back to testing the application (and see if
any application and/or back-end access issues come into play).
Hope this helps,
Robert
Charlie Therit wrote:
The short answer is that even though you have 96 threads configured, you
don't have enough runnable threads to keep the CPU(s) busy (on any tier).
Combined with the fact your WLS appears to be slow, I'd guess that your
requests are likely blocking/queueing instead of being serviced. You need
to get an idea of what your server is doing with the requests. For
example, try obtaining periodic thread dumps of the JVM to see what
activity is taking place.
In a multi-tier configuration, WLS may not be only place to investigate.
It is possible that you have poor database response time and therefore each
thread is waiting for a response. This seems plausible since you state
that you have 128 connections and therefore each thread could be blocking
on a database response. In that case, you might actually need to increase
the thread count or tune the database so that your time spent waiting is
reduced.
Since you report that all tiers are underutilized, some serious
investigation is in order. Have you monitored the network traffic?
Perhaps the amount of data being passed between tiers is significant and is
having a congestion effect? Perhaps the database server has a poor disk
layout and is spending considerable time servicing requests? Problems with
network routing between servers? As you can see there are many possible
conditions that could result in the situation you report. I've listed only
a few possibilities.
Hope this helps,
-Charlie
Madis Tapupere wrote:
Hello,
can anybody please help us with pointers on where to look in case of
the following symptoms:
We have WLS 5.1SP8 on HPUX, jvm 1.2.2.04 connecting to Oracle and
Informix
The WLS seems to be slow, even when the CPU utilisation is around 50%.
It never goes to
full 100%. It very much seems like a configuration issue, the question
just is which one...
The tiers in front and behind also report low cpu utilisation.
Current config:
max 80 concurrent requests allowed by client.
Oracle database pool 128 (I know it is too large, planning to reduce)
maxthreads 96
Weblogic oracle OCI driver
many thanks in advance
madis

Macbook pro slow with 1tb external hard drive not fully utilized

I have a 13" macbook pro (late 2011 model) 8GB ram (1333mhz) and 500gb sata drive. I use photoshop CS6 and Lightroom quite frequently and all my images are stored on a 1tb external seagate drive. The drive is 70% full but the macbook is not able to open folders. The drive does show up in the finder but seems blank though all the files are on it. I do not keep much on my internal hard drive so the computer is not inundated with data in anyway. Please let me know how I can fix this

Triple-click anywhere in the line below on this page to select it:
{ diskutil list; echo; diskutil cs list; } | open -ef
Copy the selected text to the Clipboard by pressing the key combination command-C.
Launch the Terminal application in any of the following ways:
☞ Enter the first few letters of its name into a Spotlight search. Select it in the results (it should be at the top.)
☞ In the Finder, select Go ▹ Utilities from the menu bar, or press the key combination shift-command-U. The application is in the folder that opens.
☞ Open LaunchPad. Click Utilities, then Terminal in the icon grid.
Paste into the Terminal window by pressing the key combination command-V. I've tested these instructions only with the Safari web browser. If you use another browser, you may have to press the return key after pasting.
Wait for a new line ending in a dollar sign (“$”) to appear.
A TextEdit window will open with the output of the command. If the command produced no output, the window will be empty. Post the contents of the TextEdit window (not the Terminal window), if any — the text, please, not a screenshot. The title of the window doesn't matter, and you don't need to post that.
If any personal information appears in the output, anonymize before posting, but don’t remove the context.

The new Lamb of God album does not fully download all of the songs, DO NOY BUY IT FROM HERE...T

on top of that there is not customer service here

delete the songs you have just purchased from itunes. Then, go to the store and in quick links click on "Purchased". Then find on the top right "Songs not on this computer" (or something similar) and you will find them for download. It's best to download them one at a time and wait for one to complete before moving on to the next one.

Media Encoder not fully utilizing CPU

CPU:
     - i7-4930K 3.4GHz (4.5GHz OC)
RAM:
     - 32GB Kingston 2400MHz
HDD:
     - System: Samsung 840 Evo SSD
     - Media: 2x600GB Seagate 10K SAS 6GB/s, + Highpoint Rocket RAID Controller (RAID 0)
GPU:
     - NVIDIA Quadro K2000 3GB
OS:
     - Win8.1 Pro 64Bit
I encode edits created on Premiere Pro CC 2014 to main profile H.264/12Mbps but they are not being turned around nearly as fast as I thought they would for a PC with my kind of specs, when observing the task manager during an encode I can see that the CPU utilization only goes up to a maximum of 64% (averaging at around 45-55%), meaning that a 30min sequence can take as long as 45mins to complete.
Is there a way of getting any more horse power from my computer? I have invested in this solution as i need to turn around clips very quickly but at the moment the performance is not living up to my expectations.

As far as I know, FCPX uses the GPU for rendering in the timeline now...
Not sure, if it’s of any help...?
Check out “Mark Spencer” and “Steve Martin” talking about rendering in FCPX...
http://www.fcp.co/final-cut-pro/news/610-mark-spencer-and-steve-martin-talk-gpu- rendering-in-fcpx-and-the-use-of-multiple-cores-with-compressor
Regards
Nolan

FCPX is not fully utilizing CPU. Slow Rendering.

When rendering in FCPX, I notice in the Activity Monitor that FCPX is only using 38-43% of available CPU and the rest of the CPU muscle is sitting there "idle." Does any one know how to correct this problem? This is nonsense. I should be rendering quickly and right now I am not moving faster than my 3 year old iMac with a 512 MB graphics card.
I've spent a bank roll on a top of the line 2012/13 iMac and external drive setup:
3.4 Ghz i7
32MB RAM
3 TB Fusion internal drive (only running system softward and other applications that are NOT opn when I am runing FCPX)
My FCPX files are on a Pegusus R6 12 TB Thunderbolt drive
NVIDIA GeForce GTX 680MX 2048 MB graphics card

As far as I know, FCPX uses the GPU for rendering in the timeline now...
Not sure, if it’s of any help...?
Check out “Mark Spencer” and “Steve Martin” talking about rendering in FCPX...
http://www.fcp.co/final-cut-pro/news/610-mark-spencer-and-steve-martin-talk-gpu- rendering-in-fcpx-and-the-use-of-multiple-cores-with-compressor
Regards
Nolan

How to fully use all CPUs?

Hi Folks,
I'm a newbe to the new MacPro. I have a 12 core (24 thread) system but it is only using 4 cores. I receive the message "Waiting for shared memory" How do I employ all cores? I am trying to stress test the system with Seti.
Thanks,
SD

Are you really still using 10.5.6? if so, an Upgrade to 10.6 would improve task Utilization:
Grand Central Dispatch (GCD) is a technology developed by Apple Inc. to optimize application support for systems with multi-core processors and other symmetric multiprocessing systems.[2] It is an implementation of task parallelism based on the thread pool pattern. It was first released with Mac OS X 10.6, and is also available with iOS 4 and above.
--http://en.wikipedia.org/wiki/Grand_Central_Dispatch

RAM is not fully utilizing in G430

Hi guys
yesterday i brought G430 , and i upgraded the RAM to 4gb
after installing windows xp , i checked the RAM size it showing only 2.98 GB
why it is showing this much difference
PS: i didnt install any drivers
thanks in advance

That must be because your OS is a 32 bit one.32 bit OSe's donnot have the capacity to use the entire RAM you have put in.They have limitations on how much memory it can "recognize"
Cheers and regards,
• » νιנαｿѕαяα∂нι ѕαмανє∂αм ™ « •
●๋•کáŕádhí'ک díáŕý ツ
I am a volunteer here. I don't work for Lenovo

HP Photosmart Premium All-In-One driver software will not fully install

After downloading and installing the full version of the printer driver (both from on-line site AND from purchased driver CD), the software appears to download fine, but will not fully install. I have run Print and Scan Doctor and tried on-line support, uninstalling, reinstalling the drivers and have tried on-line HP support. My new computer (HP Envy 715 xt) is less than 3 weeks old. I am running Windows 7 Professional 64 bit. The printer/scanner/copier/fax worked fine on my old computer which ran Windows XP. I have tried everything - HP on-line support will not speak to me without charging me, even though my computer is brand new - they insist the problem is with the printer, which is out of warranty.

Hello Deb!
I would recommend performing an L3 uninstall, and reinstalling.
I am listing the L3 uninstall steps below:
-Launch your temp folder
-Go to the 7z folder
-Go to the util\CCC folder
-Launch the uninstall_L3.bat file
Please let me know the results!
Please click “Accept as Solution " if you feel my post solved your issue, it will help others find the solution.
Click the “Kudos, Thumbs Up" on the right to say “Thanks" for helping!
Jamieson
I work on behalf of HP
"Remember, I'm pulling for you, we're all in this together!" - Red Green.

Oracle EPM Planning 11.1.2.1 - Ziplogs utility not working on all servers

Oracle EPM Planning 11.1.2.1 - Ziplogs utility not working on all servers
We have Planning 11.1.2.1 distributed so that Planning web app, EAS etc. are on own webservers and Essbase is on separate cluster.
Currently we have an issue with ziplogs utility - it seems to work on Essbase servers gathering all necessary ODL logs for over 100MBs, but on Planning Webservers ziplogs utility seems to start, but generates only empty 1kb zip file without any content.
We haven't touched logging otherwise, everything should be pretty much in original state.
Unfortunately I can't remember now exactly, but I have a a hunch that ziplogs utility worked prior patching Planning element on webservers to 11.1.2.1.600.
Other servers or elements haven't been patched (although Oracle recommends also FS, EB etc. to be patched to 11.1.2.1.600). I'm wondering, could this be the reason as in a way FS and EB are on same patching level whereas Planning is not?
Also, there seems to be a bit of limited diskspace on webservers (still waiting for extension), but I would imagine lack of diskspace for preparing any temp files etc. would results in some errors, the problem is that ziplogs doesn't give any errors, just runs very short time and then ends bringing 1kb empty file.
Has anyone experienced anything similar and are you perhaps able to give quick guidance how to enhance or modify ziplogs batch file etc. to make it create a bit of logging of its own actions?
Thanks in advance if anyone has ideas what could be the root cause for this or faced something similar.

Okay, I was able to get it installed and Common Components installed correctly this time. I still can't access workspace though.
Is there something I need to do before I can browse to http://localhost:19000/workspace/index.jsp?
All I get is the following error:
Error 404--Not Found
From RFC 2068 Hypertext Transfer Protocol -- HTTP/1.1:
10.4.5 404 Not Found
The server has not found anything matching the Request-URI. No indication is given of whether the condition is temporary or permanent.
If the server does not wish to make this information available to the client, the status code 403 (Forbidden) can be used instead. The 410 (Gone) status code SHOULD be used if the server knows, through some internally configurable mechanism, that an old resource is permanently unavailable and has no forwarding address.
Anyone know what's going on?

After my phone was fully submerged in water, I'm getting the overheating error message but the phone is not hot at all. Is this the result of the phone being damaged by the water?

I dropped my iPhone 4S in my friends pool. It was fully submerged for about 30 seconds.
We took it out, dried it and put it in rice. After 24 hours I checked the phone. It turns on and runs smoothly. It connects to the computer and syncs perfectly. But after a couple minutes of use I get an overheating error message. But the phone is not hot at all.
Is this the result of the phone being damaged by the water? Is there any way around this? The phone works perfectly fine and isn't overheating at all.
What can I do?

Despite the error that I get after a few minutes, the phone does work fine.
But by most peoples definition, the phone isn't working.
I just want to know if anyone has been in my boat and if the overheat message without the phone overheating can correct itself or if it goes away with sufficient drying of the phone.
I know its slim, but I figured I'd ask.

Have installed cc2014, menues are not fully visible. when changing font, system (all new sony vaio, win 8.1) crashes completely.

have installed cc2014, menues are not fully visible. when changing font, system (all new sony vaio, win 8.1) crashes completely

Dear Zak,
thank you very much! Your information solved the problem immediately!
Best regards,
Philipp Frank
New website
www.philippfrank.at<http://www.philippfrank.at>
Dipl Ing Philipp Frank
1010 Wien
Jakobergasse 4/14
+43 1 513 22 83 19
+43 650 250 5000
[email protected]<mailto:[email protected]>
www.philippfrank.at
Von: Zak Williamson (Adobe)
Gesendet: Montag, 23. Juni 2014 20:40
An: Dipl Ing Philipp Frank
Betreff: have installed cc2014, menues are not fully visible. when changing font, system (all new sony vaio, win 8.1) crashes completely.
have installed cc2014, menues are not fully visible. when changing font, system (all new sony vaio, win 8.1) crashes completely.
created by Zak Williamson (Adobe)<https://forums.adobe.com/people/ZakWilliamson%28Adobe%29> in Help with using Adobe Muse CC - View the full discussion<https://forums.adobe.com/message/6488371#6488371>

I am using final cut express. It seems that any clip over a certain length, say 5 min, does not fully transfer! All shorter clips are perfectly fine. Is there some reason why longer clips are not fully transferred? (a half blue circle shows up on the tran

I am using final cut express 4. It seems that any clip over a certain length, say 5 min, does not fully transfer! All shorter clips are perfectly fine. Is there some reason why longer clips are not fully transferred? (a half blue circle shows up on the transfer list, indicating that the clip was not fully transferred. When I check in the capture scratch folder, sure enough the whole clip is not there!
Is there some glitch in final cut express that only allows short clips to be transferred in full?

Thanks for the reply. I'll try to answer all the questions (I am a relative novice). Your help is appreciated.
By the way, the other reply I received suggested the following:
In Final Cut Express>System Settings>Scratch Discs make sure the Limit Capture/Export File Segement Size To: is not checked.
I did this and it was already unchecked. So this is not the answer...
The strange thing is, I have been logging and transferring with FCE for 2 yrs without any problems at all with shorter clips, but have come up against the problem with long clips (over 5 or so min) the whole time. I got around it by simply making sure I stopped recording after less than 5 min while shooting footage. OR if I accidentally got a long clip, I import into imovie and then import into FCE. Not very streamline and I worry about compressions altering quality!
I am now trying to import a 20 min and a 17 min clip and it only transfers around 3 minutes of each, I have re-tried several times and it happens repeatedly. The blue circle in the transfer list shows up as a half blue circle instead of full. Clip cuts off after 3 or so minutes.
ANSWERS
how you are ingesting the files
From external flash drive connected directly to mac using FCE.
the format of the source footage
digital via ext flash memory. Frame size :1929X1080 / 25fps
what camera shot the footage
SONY HXR NX5
how you are connecting to the computer
from flash memory directly via usb
what version of FCE you are using
4
what OS your computer is running
10.6.8
what model Mac
macbook Pro intel Core i7
what drive you are using for Capture Scratch
3TB external drive (plenty of room left)

[svn:bz-trunk] 14749: Change tests to use fully qualified hostname of flexteam.corp.adobe. com as flexteam was not resolving on all regression boxes.

Revision: 14749
Revision: 14749
Author:   [email protected]
Date:     2010-03-14 05:43:14 -0700 (Sun, 14 Mar 2010)
Log Message:
Change tests to use fully qualified hostname of flexteam.corp.adobe.com as flexteam was not resolving on all regression boxes.
Modified Paths:
    blazeds/trunk/qa/apps/qa-regress/testsuites/mxunit/tests/proxyService/202863/bug202863Doc Lit.mxml
    blazeds/trunk/qa/apps/qa-regress/testsuites/mxunit/tests/proxyService/202863/bug202863Rpc Enc.mxml

Oc4j not fully utilizing all cpus

Similar Messages

Maybe you are looking for