SPONSORS:






User Tag List

Results 1 to 9 of 9
  1. #1
    Member
    Join Date
    Sep 2000
    Posts
    48
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    testing IE with Silktest for a prolonged time.

    I have never been able to run any script longer than 8 hours. Internet Explorer 5.0 would just simply crash after a couple hours. I ran the same script on a win2k machine and it lasted 4 hours, but when I ran it on an nt4 machine it lasted 8 hours. On the win2k machine, IE 5.0 was closed (either by silk test or because it crashed) at the end of the test. On the nt4 machine IE5.0 just simply hung. Of course in both cases Silktest aborted the script.
    What could possibly cause IE5.0 to crash. My guess is that is an inherent instability in IE or that maybe silktest causes it to crash. It could also be a problem in our web application or my test script.
    Konrad Nau
    Software Test Automation Specialist Intern
    CQOS INC.

  2. #2
    Senior Member
    Join Date
    Jun 2000
    Location
    Chicago, IL USA
    Posts
    383
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    I would wager that this is Bill Gates' problem, not Segue's.
    Keep track of the IE memory usage through the Windows NT Task manager.
    I am sure you will see it go up and up until
    there is none left.
    SilkTest typically hangs when a memory leak hit the limit.
    I have a test that actually keeps track of the memory usage and restarts the Application when it gets critical.
    Rob Kapteyn, CSTE
    RobKapteyn@gmail.com

  3. #3
    Member
    Join Date
    Sep 2000
    Posts
    48
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    anybody ever experienced the same problem? Or is there anybody that was able to test Silktest 5.0.3.1 with IE for a 24+ hours period
    Konrad Nau
    Software Test Automation Specialist Intern
    CQOS INC.

  4. #4
    Member
    Join Date
    Dec 1999
    Location
    Seattle, WA
    Posts
    64
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    bad wager! The leak is in agent.exe. I sent a very small .T file to segue SEVERAL MONTHS AGO that clearly isolates and demonstrates the problem.

    They even told me they bought a "special tool" (bounds checker i would guess) to track down this very problem.

    I call about every month or so to follow up, but no real answers are ever provided. Reminds me, they are due another phone call! Thanks.

    ------------------
    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>It doesn't matter if I go to heaven or to hell, I have friends in both places.<HR></BLOCKQUOTE>

    [This message has been edited by Matt Sullivan (edited 09-20-2000).]
    <BLOCKQUOTE]<font size=1 face=Verdana, Arial, Helvetica]quote:</font]<HR]It doesn't matter if I go to heaven or to hell, I have friends in both places.<HR]</BLOCKQUOTE]

  5. #5
    Member
    Join Date
    Sep 2000
    Posts
    48
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    that explains why segue hasn't answered this question for me yet. I am surprised though that they haven't addressed this issue since with the memory leak problems, Silktest will be useless for stress testing.
    Konrad Nau
    Software Test Automation Specialist Intern
    CQOS INC.

  6. #6
    Member
    Join Date
    Dec 1999
    Location
    Seattle, WA
    Posts
    64
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    The conspiracy theory: Make agent leak and sell more Silk Performer!

    (this is a joke folks)

    ------------------
    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>It doesn't matter if I go to heaven or to hell, I have friends in both places.<HR></BLOCKQUOTE>

    [This message has been edited by Matt Sullivan (edited 09-21-2000).]
    <BLOCKQUOTE]<font size=1 face=Verdana, Arial, Helvetica]quote:</font]<HR]It doesn't matter if I go to heaven or to hell, I have friends in both places.<HR]</BLOCKQUOTE]

  7. #7
    Senior Member
    Join Date
    Jul 1999
    Location
    New York, NY, USA
    Posts
    137
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    Matt,

    Silk is a Lousey tool for stress testing and it wasn't designed for it. I have had no problem running Silk for 30 - 50 hours against IE looping through complex testcases (about 3k to 5k total testcase results in the file), but I think that there are certain methodes within Silk that are less stable than others. The problem is that it takes a whole lot of experience with Silk to pick and choose which methodes to use.

    If you are not running in a .PLN file you can try ocaisionally closing the browser between testcases pause for a couple of seconds and start the next testcase.

    ( Browser.Close () / Sleep (2))

    I also think that if you are running many short or fast testcases, you may be Generating such a huge results file that it may bring Silk down. I prefer to script in error handling when it is possible to continue a test past a failure.

    Tom

  8. #8
    Member
    Join Date
    Dec 1999
    Location
    Seattle, WA
    Posts
    64
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    Definitely not trying to use Silk as a stress script (twas only a joke). I completely agree on this point. Silk is best suited for cross browser functionality tests.

    On the solution of using a PLN file, note that it will not work as it does not get around the bug in agent.exe. For now, the only solution is to kill the agent.exe process between scripts. The way we do this now is with command line usage, e.g. call partner with params to run and quit the script
    <code>
    partner.exe -r -q whatever.t
    </code>
    and make suites like this...
    <code>
    partner.exe -r -q first.t
    partner.exe -r -q second.t
    partner.exe -r -q third.t
    </code>

    Also note that the browser is not problem. That's what we thought at first also, and so we already kill it between every testcase as part of DefaultTestCaseEnter override.

    Size of results files defintely aren't the problem either. Results are stored in jou files on the hard drive (not in memory) and I believe they are handled by Partner.exe, not agent.exe (which is the code executing the windows messages, and the code in which we see the slowdown)

    Most importantly here, the repro which I sent to Segue very clearly points to Agent.exe, and last week they told me the developer builds now have a correction to the code. It is to be released in version 5.5, scheduled for December.

    Hope this helps.

    ------------------
    <BLOCKQUOTE><font size="1" face="Verdana, Arial, Helvetica">quote:</font><HR>It doesn't matter if I go to heaven or to hell, I have friends in both places.<HR></BLOCKQUOTE>
    <BLOCKQUOTE]<font size=1 face=Verdana, Arial, Helvetica]quote:</font]<HR]It doesn't matter if I go to heaven or to hell, I have friends in both places.<HR]</BLOCKQUOTE]

  9. #9
    Senior Member
    Join Date
    Jul 1999
    Location
    New York, NY, USA
    Posts
    137
    Post Thanks / Like
    Mentioned
    0 Post(s)
    Tagged
    0 Thread(s)

    Re: testing IE with Silktest for a prolonged time.

    Matt,
    Well here's hoping that Segue finaly has a "Fix". I have had similar problems, but have been able to get by with some of the solutions I listed. I wasn't trying to imply that anyone was inexperienced with Silk, rather that we all seem to have different problems related to the same bug. I gave up on running prolonged testcases against Netscape because no matter what I did the agent seemed to leak memory so bad that I would crash. The problem was that sometimes the crash was in the agaent, sometimes the browser would just hange and Silk would not hang as well and sometimes I would get a visit from theold Dr. (Watson).

    Please let me know if anyone finds a real solution to this.

    Thanks,
    Tom

 

 

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

vBulletin Optimisation provided by vB Optimise v2.6.0 Beta 4 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging v3.0.9 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Questions / Answers Form provided by vBAnswers (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
vBNominatevBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Feedback Buttons provided by Advanced Post Thanks / Like (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Username Changing provided by Username Change (Free) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
BetaSoft Inc.
Digital Point modules: Sphinx-based search
All times are GMT -8. The time now is 03:59 AM.

Copyright BetaSoft Inc.