For Andrè
If you read in a better way my mail, you find the attachment
*wsloader.txt *that
is the code of the bad webapp :D
For Chuck
Is Java really cross-platform?
For all:
I forgot to tell you that we are using the Strong Cryptography Library "
Java Cryptography Extension (JCE) Unlimited Strength Jurisdiction Policy
Files 5.0" and the webservices that are on the other side run unde IBM
websphere.
For linux:
I don't want to migrate the application on windowz (my boss choice)
For all:
Be patient for (if/any) my english mistakes

Michele

On Fri, Sep 3, 2010 at 3:11 PM, André Warnier <a...@ice-sa.com> wrote:

> Michele,
>
> I just want to clarify my previous answers :
>
> I am not saying that the problem that you encounter is *necessarily* a bug
> in one or the other JVM.
> You have not shown the code of your webapp, so we cannot tell you that the
> problem is there either.
> One thing we can tell you, is that the problem, as you show it, is *not* in
> the Tomcat code, because in this case Tomcat does not participate in what
> your webapp is doing.  It is your webapp, and your webapp alone, which sets
> up this SSL connection with another server.  The Tomcat code is not involved
> in this.
> Without a very detailed examination of your webapp code, it is not possible
> to say if the problem is there, and I doubt that anyone here really has the
> time or inclination to study your webapp code in detail.
> But on the surface, I would say that the chances are at least 10:1 that it
> is your webapp code that generates the problem.  It may be doing something
> unsafe, which for some reason to do with the different platforms, or
> differences in the JVMs between those platforms, or differences between the
> loads on these platforms, shows up under Linux and not under Windows.
> So the recommendation to try a more recent JVM is not because I am sure
> that it is 100% of the solution to your problem.  But it is the "cheapest"
> way to find out if by changing the environment a little bit, the problem
> still appears in the same way, or not.
>
> Under both Windows and Linux, you should be able to install a 1.6 JVM next
> to the 1.5 JVM, and then just set the JAVA_HOME environment variable of your
> Tomcat, to run under the one or the other.
>
> Maybe if you upgrade both the Windows and the Linux JVM to 1.6.21, then the
> problem will start appearing on both platforms.  That would be a clearer
> sign that the problem is in the webapp.
> And if the problem then still appears only under Linux, then it is worth
> looking deeper.
> If the problem does not appear at all anymore, then there are 2
> possibilities :
> - the problem was due to a bug in the Linux 1.5 JVM or one of the
> underlying native libraries
> OR
> - the problem is still in the webapp, but it does not show up so easily
> anymore under the 1.6 JVM (that would be the worse outcome, because then you
> are not sure anymore when it will hit you again)
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> André Warnier wrote:
>
>> Michele Mase' wrote:
>>
>>> Both windoz and linux use the same java :(
>>>
>>
>> Well no, they are NOT the same, even if they have the same version.
>> The "Windows java JVM" is a Windows executable program (java.exe).  The
>> Linux java JVM is a Linux executable program.  Each is compiled from
>> presumably much the same code, but there are significant differences between
>> them, such as for example the fact that they are linked to different native
>> libraries (DLLs under Windows, .so libraries under Unix/Linux).
>>
>> So a bug can exist in one, and not in the other.
>> And code that has to do with TCP/IP and SSL is likely quite different
>> under each platform.
>>
>> Anyway, java 1.5.0 is quite old.
>>  From the java website :
>>
>> J2SE 5.0 End of Service Life Notice
>> J2SE 5.0 reached its End of Service Life (EOSL) on November 3, 2009, which
>> is the date of the final publicly available update of version 5.0 (J2SE 5.0
>> Update 22).
>>
>> You should try a recent 1.6 release (1.6.21 ?), and see if the problem
>> still exists.
>> Or you can continue posting the same question on this forum every couple
>> of months, but it is unlikely that anyone would be very interested in
>> investigating much further, until you try it and report that the issue
>> appears in a recent JVM too.
>>
>>
>>
>>> On Fri, Sep 3, 2010 at 12:01 PM, André Warnier <a...@ice-sa.com> wrote:
>>>
>>>  Hi.
>>>>
>>>> From what I can see below (and what you explain yourself), this problem
>>>> has
>>>> nothing to do with Tomcat.  It is the (your) webapp which uses an SSL
>>>> connection to some other server, and which receives this exception.
>>>>  Tomcat
>>>> does not even know that this is happening.
>>>> Tomcat in this case is just the engine which runs your webapp.
>>>>
>>>> What may have something to do with the error, is the java JVM which is
>>>> used
>>>> to run both Tomcat and your webapp.  Have you tried updating the JVM to
>>>> a
>>>> more recent version (like 1.6) ? You can run the same Tomcat (and your
>>>> webapp) under the newer JVM without any change.
>>>>
>>>> The reason why it happens under Linux and not under Windows, is most
>>>> probably because the JVM is different.
>>>>
>>>>
>>>> Michele Mase' wrote:
>>>>
>>>>  On Sun, Jul 25, 2010 at 9:06 PM, Michele Mase' <michele.m...@gmail.com
>>>>>
>>>>>> wrote:
>>>>>>
>>>>>  Hi folks!
>>>>>
>>>>>> I've a strange problem, please help me to find a solution (not telling
>>>>>> me
>>>>>> to make a script in order restart tomcat in case of the exception)
>>>>>> Under linux environment,
>>>>>> RedHat EL5.5
>>>>>> Jdk 1.5.0_22
>>>>>> Tomcat6.0.26
>>>>>> axis1.4
>>>>>> our webapps takes strange intermittent "javax.net.ssl.SSLException:
>>>>>> Invalid
>>>>>> padding" errors.
>>>>>> The same webapp under a windows system never catches the exception
>>>>>> The webapps uses tomcat like a client with the axis library (1.4
>>>>>> version
>>>>>> only, it is non axis <1.4 capable) in order to connect to an external
>>>>>> webservice with https.
>>>>>> You catch the exception after 1 hour of work, 5, 7 hours and more than
>>>>>> 24
>>>>>> hours of work.
>>>>>> Once the exception is catched, the only solution to make the webapp
>>>>>> can
>>>>>> work again, is to restart the tomcat.
>>>>>> Under the windows machine we never caught the exception.
>>>>>>
>>>>>> Attachments:
>>>>>> wsloader.txt is the code for the invocation of webservices.
>>>>>> I also attach the wireshark compatible files of both situations:
>>>>>> interop91ko.enc when it doesn't work; you can take a look at the
>>>>>> pattern
>>>>>> 294.  http://www.ietf.org/rfc/rfc2246.txt?number=2246
>>>>>>
>>>>>> bad_record_mac
>>>>>>      This alert is returned if a record is received with an incorrect
>>>>>>      MAC. This message is always fatal.
>>>>>>
>>>>>> interop91ok.enc when it work
>>>>>>
>>>>>> The attachment logs-ko.txt is the application log when it doesn't work
>>>>>> The attachment logs-ok.txt is the application log when it works.
>>>>>>
>>>>>> pls, Help me! My boss wants to use the webapp under windows (migrating
>>>>>> it
>>>>>> from linux) since in windoz test environment we have never seen the
>>>>>> exception
>>>>>>
>>>>>> Regards Michele Masč
>>>>>>
>>>>>>
>>>>>>
>>>>>> ------------------------------------------------------------------------
>>>>>
>>>>>
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
>>>>> For additional commands, e-mail: users-h...@tomcat.apache.org
>>>>>
>>>>>
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
>>>> For additional commands, e-mail: users-h...@tomcat.apache.org
>>>>
>>>>
>>>>
>>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
>> For additional commands, e-mail: users-h...@tomcat.apache.org
>>
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
> For additional commands, e-mail: users-h...@tomcat.apache.org
>
>

Reply via email to