« Feature Requests for Adobe Reader | Main | Issues related to Printing and Uninstallation of Reader 8.1.2 on Windows »

100% CPU Utilization Problem

Some of our users have reported a 100% CPU utilization problem while using Adobe Reader 8.x for Unix platforms. This has been discussed on the user-to-user forums post but unfortunately we are unable to replicate this issue locally. Some of the users have reported that this problem seems to go away if you set the proxy in Preferences->Internet to “Manual Proxy Configuration” and enter the proxy server and associated port #, then exit Acroread and restart.

I would request all those users who have faced this issue to post comments on how they reproduced the problem and give some information on their OS/distribution, etc. to help us isolate the issue. Thanks for your continuous feedback and comments!


    I had this issue on a 2-core machine. Starting one Acrobat (in Firefox) would hog one core 100%. Starting another instance would hog the other to 100%. This is on SuSE Linux Enterprise 10.1.

    I tried the proxy fix and it worked. I’m behind a corporate intranet so the proxy is a plausible cause for this (how I don’t know, that’s your task now, Adobe).

    Thanks for the tip!

    I can replicate this issue quite easily when using Reader as a Firefox plugin, just by surfing to random PDF files on the Internet. Sometimes the problem manifests and sometimes it does not. CPU usage stays constantly high on one core (though not 100%) and memory use increases until the acroread process is killed.

    I have a Direct connection (no proxy) configured.

    System details:

    Fedora 8, Quad-Core
    Firefox x86_64
    Adobe Reader 8.1.2 rpm


    You need look at the acrobat code that is trying to access pmupsw.adobe.com. Acroread gets into an infinite loop when it can’t access this site. I ran wireshark and it chokes right after requesting this site ip from dns.

    sorry for cross commenting, i found this after posting comment to “Known Issues with Reader 8.1.2” i think this is correct place.

    We have this issues 100% CPU usage and up to 200MB memory usage for 100Kb pdf dokument:

    on Windows Systems with AR8.1.2

    We have some converted pdf dokuments (from eps), which causes a high cpu usage on Windows XP/Vista systems. Same dokument causes AR8.1.2 crashes on other Windows XP Systems. Same documents can be opened with AR7.x

    We found that this issue comes while using Adobe Distiller 7.0.9 or ghostscript (with multiple devices setting) but not with Adobe Distiller 8.1

    Can you please point out the changes/issues from Distiller 7.0.9 to 8.1 an Reader 7.x to 8.x

    thanks in advance

    I have the same problem here on two machines, both fedora 8 on core 2 duos (e6600, 7100). I have to kill it using signal 9 to get rid of it. I managed to reproduce it also on centos 5.1 (redhat enterprise 5.1). Acroread is started as a firefox plugin and stays in the process list even after finishing firefox. It is really annoying and I am close to remove acroread if this problem is not getting fixed soon. I do not see why Adobe cannot reproduce the problem. Get a maschine up and running (use xen or kvm for testing if you bother setting up a system freshly) on fedora 8, install firefox and your acroread by yum. That is done in less than 15 minutes.

    I am seeing this issue on a machine running Fedora Core 8 with Acrobat 8.1.2

    As far as I can tell it’s a memory leak. The problem occurs only when a pdf is open from within the browser. While that document is open, everything is fine. As soon as I close it, I see the memory useage by the process start edging up by about 0.2% per second. Eventually, if left unattended, it consumes all available memory and crashes my computer. This is pretty impressive since this is the only program that ever compromises my linux box stability.

    I believe, if I just open a pdf file that is already on my hard drive, this doesn’t happen.

    I’m seeing the same symptoms with Acrobat reader 8.1.2 on RHEL 5.2. When acroread is started as a Firefox plugin its cpu utilisation is very high. This usually occurs after the document is closed in the browser. I also notice that the xorg process is forced to use a lot of cpu whilst this is happening, this drops back to normal once the acroread process is killed.

    Another issue is that the search features don’t work in the plugin, you can’t type in the search box.

    I am having the same problem when I open PDF using firefox. I am using Kubuntu Hardy.

    I experience the same problem (100% utilization with ld-linux.so.2 process – it keeps running after using acrobat reader plugin in firefox – must be killed manually), I am running Gentoo Linux. Here are the details of my system:

    * Gentoo x86/64 (64 bit system)
    * Kernel 2.6.24
    * KDE 3.5.9
    * Acrobat Reader 8.1.2
    * Firefox

    I also tried to disable “fast web view” as suggested by some users, without any improvement.

    It would be really helpful if you managed to replicate and solve the problem.


    RHEL5 and acroread version 101:8.1.2-1

    Will stay resident consuming 50% of my CPU (1 core) after being used in firefox as a plugin.

    BTW can you please trim down the reader ? 80 Megs is a little too extravagant. I agree you would like to push in forms etc but there will be a mass migration to alternatives if you insist on packaging pork barrels.

    openSuSE 10.3 KDE
    firefox 3.0.1-2.1

    yeah this is getting really annoying. I don’t see this happen if I view the pdf with regular adobe acrobat reader (only happens when I view in firefox).

    It doesn’t always happen immediately (but can).

    If I close the web page (window or tab) it doesn’t end the ‘acroread’ PID). I have to kill it. Sometimes it won’t start “hogging” the CPU for a few hours after the process has been running, or even after the window/browser has been closed.


    In the Windows version you can disable the auto-update feature, but I can’t find the same option in the Linux version. Am I missing this or it is not there?


    The Problem still exists. Even on Fedora 9 and Centos 5.2 (core2duo, 2,4MHz, 3GB RAM). I cannot understand why Adobe is not able to solve this problem. Have a look at the output of “top” (rearranged for the forum):


    4088 797m 56 21.9

    7:26.09 acroread

    3425 135m 44 2.4
    9:20.05 X

    RedHat Enterprise 5.2
    Dell Precision 360 – dual processor

    The same for me, on Ubuntu blabla…
    Sounds like a memory leak in the application.

    I can reproduce the problem every time on any machine running centos 5.2.

    The problem happens if you open a pdf from inside firefox. When you close the tab with the pdf in it the cpu will spike.

    I tried this with Opera instead of firefox and the problem does not happen

    The correct proxy settings fix the bug for me too.
    What to say else that I’m not impressed by adobe developpers…
    A simple googling shows that everybody seem to have the bug, except adobe: “unfortunately we are unable to replicate this issue locally”


    I also had this problem, but now it seems to get solved 🙂

    My config:
    Amd turion x2 notebook (fujitsu-siemens).
    OS: Debian testing release (lenny), 32 bit.
    I installed the reader from the deb package.

    Im not sure, that it was the thing that solved, but after the kernel upgrade from 2.6.26-1 to 2.6.28 I installed the reader again, and now it works well as standalone, and also in iceweasel/firefox (i had to compile the kernel from source).

    There was also another change in the environment, I installed the latest official Nvidia Geforce driver on my notebook. Before that I was using the free driver shipped with Debian.

    I hope this helps You too!

    I got some news Everyone!

    The system is still Debian “lenny” on AMD Turion X2 notebook, linux-kernel 2.6.28.

    I installed the compiz, and after I started it, the 100% CPU utilization problem came back!

    After I uninstalled compiz, the problem disappeared again!


    The solution worked for me as well. I’m using Fedora 7 with Acroread 8.1.3. Without setting the proxy, Acroread would utilize 100% of the CPU… which is what lead me to this web page. I am behind a corporate firewall and proxy, so once I manually set the proxy and restarted Acroread, my processor utilization was back to normal. Thanks!

Sorry, the comment form is closed at this time.

Copyright © 2017 Adobe Systems Incorporated. All rights reserved.
Terms of Use | Privacy Policy and Cookies (Updated)