wrapper-user — User issues for Wrapper. (Support questions, etc.)

[Wrapper-user] Timeout properties appear to be ignored

[Matthew V. <Mat...@EM...>]

Hi,

=20

I have recently upgraded to version 3.1.2 (from 2.2.3 I think) and I've
run into a confusing problem. Despite the presence of the following
properties in the .conf file, timeouts are occurring. The timeouts take
approximately 5 minutes and iterate 5 times; on Unix-flavored machines
this is a medium-sized problem, on Windows, this blocks Windows startup,
adding 30 minutes to the blank screen before your typical user sees his
login dialog.

=20

Relevant properties:

=20

wrapper.startup.timeout=3D0

wrapper.ping.timeout=3D0

=20

Log section:

=20

ERROR  | wrapper  | 2005/05/03 14:52:22 | Startup failed: Timed out
waiting for signal from JVM.

ERROR  | wrapper  | 2005/05/03 14:52:22 | JVM did not exit on request,
terminated

STATUS | wrapper  | 2005/05/03 14:52:27 | Launching a JVM...

INFO   | jvm 2    | 2005/05/03 14:52:30 | Wrapper (Version 3.1.2)
http://wrapper.tanukisoftware.org <http://wrapper.tanukisoftware.org/>=20

INFO   | jvm 2    | 2005/05/03 14:52:30 |=20

=20

I should mention that the extra information from setting
wrapper.debug=3Dtrue hasn't provided any clues. I've also tried =
switching
the JVM (Sun/Jrockit) to no gain. I have double and triple-checked the
formatting of the configuration file, as well as setting all of the
timeout-related properties to 0 (disabled). I haven't seen any similar
bugs/issues posted, so I'm a bit concerned.

=20

If anyone has any suggestions/solutions, I would be forever in your
debt.

=20

Thanks,

Matt

Re: [Wrapper-user] New user question

[Bashiro <ba...@en...>]

Check this link:
http://bdn.borland.com/article/0,1410,32068,00.html

it's been explained in details with examples.
That's what I followed to get my application working

bashiro


> I have a question,
> All the documentation seems to suggest that I need to define the starting
> class manually in the config, However I am just trying to run a packaged
> jar
> file as a service, the Jar files contain a manifest with the correct
> classdef in it.
>
> on a command line I need to just give the command
> java -jar rules.jar -start
>
> How can I do this using wrapper or will I have to fiddle with the config
> and
> put the full path to the main in the Wrapper config.
>
> (also I have seen nothing about running multiple services is this going to
> be a problem - how do I make sure the config files are not interfereing
> with
> it? - do I need to use the -c switch to do each service manually?
>
> Mike.
> --
> No virus found in this outgoing message.
> Checked by AVG Anti-Virus.
> Version: 7.0.308 / Virus Database: 266.11.5 - Release Date: 04/05/2005
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by: NEC IT Guy Games.
> Get your fingers limbered up and give it your best shot. 4 great events, 4
> opportunities to win big! Highest score wins.NEC IT Guy Games. Play to
> win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20
> _______________________________________________
> Wrapper-user mailing list
> Wra...@li...
> https://lists.sourceforge.net/lists/listinfo/wrapper-user
>

[Wrapper-user] New user question

[Mike <mi...@m-...>]

I have a question,
All the documentation seems to suggest that I need to define the starting
class manually in the config, However I am just trying to run a packaged jar
file as a service, the Jar files contain a manifest with the correct
classdef in it.

on a command line I need to just give the command
java -jar rules.jar -start

How can I do this using wrapper or will I have to fiddle with the config and
put the full path to the main in the Wrapper config.

(also I have seen nothing about running multiple services is this going to
be a problem - how do I make sure the config files are not interfereing with
it? - do I need to use the -c switch to do each service manually?

Mike.
--
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.308 / Virus Database: 266.11.5 - Release Date: 04/05/2005

[Wrapper-user] wrapper.filter.action=restart issue.

[Anil N. J. <aju...@ho...>]

Hi,
The JSW is working just fine as of now for my application needs and am =
very happy with it. Currently, I am trying to further enhance my product =
features using the wrapper.filter.action property. I am trying to =
restart the 'server' of my application which works fine using this =
property. However, there are a few other applications that are dependent =
on the 'server' which do not shut down when the 'server' of my =
application stops. Also, I do not see the 'wrapper.pid' file being =
created again even though the restart works fine and the application =
starts again without any issues.
Do I have to set any other properties when using the =
wrapper.filter.trigger - action properties to stop all the dependent =
services? Please advice....
Thanks,
----AJ

Re: [Wrapper-user] Application Stopped Unexpected and amazingly

[jawad b. <bok...@ya...>]

Thanks for your valuable information. 

Regards,

jawad

--- Leif Mortenson <le...@ta...> wrote:

> Jawad,
>     The timing bug is well understood and I am 100%
> sure that it has 
> been fixed.
> Versions earlier to 3.1.0 were all fine, but in
> 3.1.0 I reworked the 
> internal timing
> a bit so I could start experimenting with the new
> Tick based timer that 
> will be
> standard in 3.2.0.   When I did this, I made what I
> thought was an innocuous
> change.  The tick counters use a 32 bit integer to
> count the ticks.  It 
> is expected
> to roll over and is set up to happen 10 minutes
> after starting so it is 
> well tested
> through use.
> 
>     The problem was that when using the old system
> based timer, the code 
> which
> converted the system time into a tick count had a
> bug that would cause a 
> huge
> jump in the tick count at very specific times.  
> This would look like 
> the Wrapper
> or JVM had been hung for several hours and it would
> trigger all kinds of 
> timeouts.
> 
>     That conversion has now been fixed in 3.1.2 and
> has been working great.
> 
> Cheers,
> Leif
> 
>    
> 
> jawad bokhari wrote:
> 
> >Dear Leif,
> >
> >Yes, I was using version 3.1.0. 
> >I didn't review updates on the site for newer
> version
> >because this old version was even working perfectly
> >for me.
> >
> >Can I find some more information about this timing
> >bug?
> >
> >However, I'll upgrade to version 3.1.2 now and see
> if
> >it happens again.
> >
> >I set log-level to ERROR because I am using my own
> >written logging application similar to log4j and
> that
> >is configured my application internally.
> >If application crashes due to any reason, my
> logging
> >module informs me of that.
> >
> >But through logs I noticed that application was
> >stopped  normally and application didn't notice any
> >severity inside. It just received a stop signal
> just
> >like a normal application-stop.
> >
> >Thanks a lot for your quick attention and I'll
> apply
> >the latest updates now.
> >
> >I would appreciate if you can give me some more
> >information on timing-bug problem that caused this.
> >
> >
> >Regards,
> >
> >Jawad
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >--- Leif Mortenson <le...@ta...> wrote:
> >  
> >
> >>Jawad,
> >>    What version are you using? The only thing
> that
> >>I am aware of that could
> >>have possibly caused that is a timing bug in bug
> in
> >>3.1.0 and 3.1.1.  It was
> >>fixed in 3.1.2.  From the release notes:
> >>
> >>* Fix a problem where the JVM would restart at
> >>certain times when using the
> >>  system time based timer due to an overflow
> error. 
> >>This problem was
> >>  introduced in 3.1.0.  Due to a separate bug in
> >>3.1.0, the Wrapper would
> >>  shutdown rather than simply restarting the JVM
> as
> >>was happening in 3.1.1.
> >>  The last restart happened on Aug 21, 2004.  It
> >>will next occur Oct 10, 
> >>2004
> >>  and repeat at regular intervals.  There are no
> >>problems when using the new
> >>  Tick based timer.  Bug #1014405.
> >>
> >>    Also, what is the reason that you are setting
> >>your log level to ERROR?
> >>The way I designed the Wrapper, that is in general
> a
> >>bad idea as it 
> >>hides all
> >>of the output that you will need to recover from a
> >>problem.   If the problem
> >>is that your Java application is sending too much
> >>output to its console, 
> >>then
> >>I would suggest using a Java side logging system
> >>like log4j or LogKit to
> >>reduce or redirect that output.
> >>    Depending on your answer here, I may be able
> to
> >>come up with a change
> >>to the Wrapper to resolve whatever problem you had
> >>been having as well.
> >>
> >>Cheers,
> >>Leif
> >>
> >>jawad bokhari wrote:
> >>
> >>    
> >>
> >>>I recently noticed an amazingly strange problem
> in
> >>>using wrapper for my applicaiton.
> >>>
> >>>Application stopped working at 5 different
> servers
> >>>      
> >>>
> >>at
> >>    
> >>
> >>>sametime. 
> >>>I couldnt' see any wrapper.log having my log
> level
> >>>      
> >>>
> >>set
> >>    
> >>
> >>>to ERROR. I dont' see any error in my
> application's
> >>>log.
> >>>
> >>>But it suddenly happenend so. 
> >>>I wonder if there's some ping like application
> that
> >>>keep monitoring the service and stops the service
> >>>externally.
> >>>
> >>>Anything I should investigate in wrapper
> >>>configurations?
> >>>
> >>>
> >>>Regards,
> >>>
> >>>Jawad
> >>> 
> >>>
> >>>      
> >>>
> >>
> >>
> >>    
> >>
>
>-------------------------------------------------------
> >  
> >
> >>SF.Net email is sponsored by: Tell us your
> software
> >>development plans!
> >>Take this survey and enter to win a one-year sub
> to
> >>SourceForge.net
> >>Plus IDC's 2005 look-ahead and a copy of this
> survey
> >>Click here to start! 
> >>http://www.idcswdc.com/cgi-bin/survey?id=105hix
> >>_______________________________________________
> >>Wrapper-user mailing list
> >>Wra...@li...
> >>
> >>    
> >>
>
>https://lists.sourceforge.net/lists/listinfo/wrapper-user
> >  
> >
> >
> >__________________________________________________
> >Do You Yahoo!?
> >Tired of spam?  Yahoo! Mail has the best spam
> protection around 
> >http://mail.yahoo.com 
> >
> >
>
>-------------------------------------------------------
> >SF.Net email is sponsored by: Tell us your software
> development plans!
> >Take this survey and enter to win a one-year sub to
> SourceForge.net
> >Plus IDC's 2005 look-ahead and a copy of this
> survey
> >Click here to start! 
> http://www.idcswdc.com/cgi-bin/survey?id=105hix
> >_______________________________________________
> >Wrapper-user mailing list
> >Wra...@li...
>
>https://lists.sourceforge.net/lists/listinfo/wrapper-user
> >
> >  
> >
> 
> 
> 
>
-------------------------------------------------------
> SF.Net email is sponsored by: Tell us your software
> development plans!
> Take this survey and enter to win a one-year sub to
> SourceForge.net
> Plus IDC's 2005 look-ahead and a copy of this survey
> Click here to start! 
> http://www.idcswdc.com/cgi-bin/survey?id=105hix
> _______________________________________________
> Wrapper-user mailing list
> Wra...@li...
>
https://lists.sourceforge.net/lists/listinfo/wrapper-user
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

RE: [Wrapper-user] Wrapper under Solaris 64 bits

[Alexandre M. <am...@in...>]

Thanks, the workaround works just fine.
I've done some attempts to compile the 64 bits version, with no success yet.
(I'll have to upgrade gcc first). I'll let you know when I have made
progress with that.
Thanks for the quick answer, and thanks for this great piece of software.
Cheers,
-Alex

-----Original Message-----
From: Leif Mortenson [mailto:le...@ta...] 
Sent: mercredi 27 avril 2005 17:58
To: wra...@li...
Subject: Re: [Wrapper-user] Wrapper under Solaris 64 bits

Alex,

Alexandre Morin wrote:

> Hi, I'm using the service wrapper under Solaris 9 (64 bits).
>
> First, I get this warning which, I guess is because the .so is 32bits (?)
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | WARNING - Unable to load the
> Wrapper's native library 'libwrapper.so'.
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | The file is located on the path
> at the following location but
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | could not be loaded:
>
> INFO | jvm 1 | 2005/03/23 17:38:03 |
> /opt/InfoVista/VistaMart/wrapper/libwrapper.so
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | Please verify that the file is
> readable by the current user
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | and that the file has not been
> corrupted in any way.
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | System signals will not be
> handled correctly.
>
As you guessed, this is being caused because you are running a 64-bit JVM
with a 32-bit JNI library. You will need to compile a 64-bit version of the
Wrapper and use that. The wrapper binary should be fine as 32-bit as it is
a separate process, but the library needs to match the JVM.

If you have the experience, I would appreciate it if you could try modifying
the Makefile.solaris file that ships with the source and coming up with a
Makefile.solaris64 which I can include in future releases of the Wrapper.

> But the problem is that it seem that it is not possible to use more
> than 4Gb of memory (values above get truncated down to 4Gb). Is there
> a workaround to go beyond the 4Gb limit?
>
The Wrapper is actually doing some error checking on the properties
right now.
If you use the wrapper.java.maxmemory property to set the memory then it is
being forced to a range of 1..4096 MB. I need to be checking whether the
version of the Wrapper is 32 or 64 bits. If it is the later, this can
obviously
be much larger.

For now, if you are using 3.1.0 or later, you can set the memory settings
as you would normally by omitting the wrapper.java.maxmemory property
and then using the following instead.
wrapper.java.additional.1=-Xmx6000m

Alternatively, if you are building from source (as above) Then you may just
want to increase the maximum range. Search for 4096 in the wrapper.c
source file. There are 2 places where the upper limit is being forced to
4096.

I don't have a system that I an try this on, so post back with the results.

Cheers,
Leif



-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans!
Take this survey and enter to win a one-year sub to SourceForge.net
Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start!  http://www.idcswdc.com/cgi-bin/survey?id=105hix
_______________________________________________
Wrapper-user mailing list
Wra...@li...
https://lists.sourceforge.net/lists/listinfo/wrapper-user

Re: [Wrapper-user] Application Stopped Unexpected and amazingly

[Leif M. <le...@ta...>]

Jawad,
    The timing bug is well understood and I am 100% sure that it has 
been fixed.
Versions earlier to 3.1.0 were all fine, but in 3.1.0 I reworked the 
internal timing
a bit so I could start experimenting with the new Tick based timer that 
will be
standard in 3.2.0.   When I did this, I made what I thought was an innocuous
change.  The tick counters use a 32 bit integer to count the ticks.  It 
is expected
to roll over and is set up to happen 10 minutes after starting so it is 
well tested
through use.

    The problem was that when using the old system based timer, the code 
which
converted the system time into a tick count had a bug that would cause a 
huge
jump in the tick count at very specific times.   This would look like 
the Wrapper
or JVM had been hung for several hours and it would trigger all kinds of 
timeouts.

    That conversion has now been fixed in 3.1.2 and has been working great.

Cheers,
Leif

   

jawad bokhari wrote:

>Dear Leif,
>
>Yes, I was using version 3.1.0. 
>I didn't review updates on the site for newer version
>because this old version was even working perfectly
>for me.
>
>Can I find some more information about this timing
>bug?
>
>However, I'll upgrade to version 3.1.2 now and see if
>it happens again.
>
>I set log-level to ERROR because I am using my own
>written logging application similar to log4j and that
>is configured my application internally.
>If application crashes due to any reason, my logging
>module informs me of that.
>
>But through logs I noticed that application was
>stopped  normally and application didn't notice any
>severity inside. It just received a stop signal just
>like a normal application-stop.
>
>Thanks a lot for your quick attention and I'll apply
>the latest updates now.
>
>I would appreciate if you can give me some more
>information on timing-bug problem that caused this.
>
>
>Regards,
>
>Jawad
>
>
>
>
>
>
>
>
>
>
>--- Leif Mortenson <le...@ta...> wrote:
>  
>
>>Jawad,
>>    What version are you using? The only thing that
>>I am aware of that could
>>have possibly caused that is a timing bug in bug in
>>3.1.0 and 3.1.1.  It was
>>fixed in 3.1.2.  From the release notes:
>>
>>* Fix a problem where the JVM would restart at
>>certain times when using the
>>  system time based timer due to an overflow error. 
>>This problem was
>>  introduced in 3.1.0.  Due to a separate bug in
>>3.1.0, the Wrapper would
>>  shutdown rather than simply restarting the JVM as
>>was happening in 3.1.1.
>>  The last restart happened on Aug 21, 2004.  It
>>will next occur Oct 10, 
>>2004
>>  and repeat at regular intervals.  There are no
>>problems when using the new
>>  Tick based timer.  Bug #1014405.
>>
>>    Also, what is the reason that you are setting
>>your log level to ERROR?
>>The way I designed the Wrapper, that is in general a
>>bad idea as it 
>>hides all
>>of the output that you will need to recover from a
>>problem.   If the problem
>>is that your Java application is sending too much
>>output to its console, 
>>then
>>I would suggest using a Java side logging system
>>like log4j or LogKit to
>>reduce or redirect that output.
>>    Depending on your answer here, I may be able to
>>come up with a change
>>to the Wrapper to resolve whatever problem you had
>>been having as well.
>>
>>Cheers,
>>Leif
>>
>>jawad bokhari wrote:
>>
>>    
>>
>>>I recently noticed an amazingly strange problem in
>>>using wrapper for my applicaiton.
>>>
>>>Application stopped working at 5 different servers
>>>      
>>>
>>at
>>    
>>
>>>sametime. 
>>>I couldnt' see any wrapper.log having my log level
>>>      
>>>
>>set
>>    
>>
>>>to ERROR. I dont' see any error in my application's
>>>log.
>>>
>>>But it suddenly happenend so. 
>>>I wonder if there's some ping like application that
>>>keep monitoring the service and stops the service
>>>externally.
>>>
>>>Anything I should investigate in wrapper
>>>configurations?
>>>
>>>
>>>Regards,
>>>
>>>Jawad
>>> 
>>>
>>>      
>>>
>>
>>
>>    
>>
>-------------------------------------------------------
>  
>
>>SF.Net email is sponsored by: Tell us your software
>>development plans!
>>Take this survey and enter to win a one-year sub to
>>SourceForge.net
>>Plus IDC's 2005 look-ahead and a copy of this survey
>>Click here to start! 
>>http://www.idcswdc.com/cgi-bin/survey?id=105hix
>>_______________________________________________
>>Wrapper-user mailing list
>>Wra...@li...
>>
>>    
>>
>https://lists.sourceforge.net/lists/listinfo/wrapper-user
>  
>
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam?  Yahoo! Mail has the best spam protection around 
>http://mail.yahoo.com 
>
>
>-------------------------------------------------------
>SF.Net email is sponsored by: Tell us your software development plans!
>Take this survey and enter to win a one-year sub to SourceForge.net
>Plus IDC's 2005 look-ahead and a copy of this survey
>Click here to start!  http://www.idcswdc.com/cgi-bin/survey?id=105hix
>_______________________________________________
>Wrapper-user mailing list
>Wra...@li...
>https://lists.sourceforge.net/lists/listinfo/wrapper-user
>
>  
>

[Wrapper-user] version 3.2.0

[<do...@my...>]

When is it going to go public ?

(We anxiety wait for the System.in feature ... ;-)



Thanks

_______________________________________________
No banners. No pop-ups. No kidding.
Make My Way your home on the Web - http://www.myway.com

Re: [Wrapper-user] Application Stopped Unexpected and amazingly

[jawad b. <bok...@ya...>]

Dear Leif,

Yes, I was using version 3.1.0. 
I didn't review updates on the site for newer version
because this old version was even working perfectly
for me.

Can I find some more information about this timing
bug?

However, I'll upgrade to version 3.1.2 now and see if
it happens again.

I set log-level to ERROR because I am using my own
written logging application similar to log4j and that
is configured my application internally.
If application crashes due to any reason, my logging
module informs me of that.

But through logs I noticed that application was
stopped  normally and application didn't notice any
severity inside. It just received a stop signal just
like a normal application-stop.

Thanks a lot for your quick attention and I'll apply
the latest updates now.

I would appreciate if you can give me some more
information on timing-bug problem that caused this.


Regards,

Jawad










--- Leif Mortenson <le...@ta...> wrote:
> Jawad,
>     What version are you using? The only thing that
> I am aware of that could
> have possibly caused that is a timing bug in bug in
> 3.1.0 and 3.1.1.  It was
> fixed in 3.1.2.  From the release notes:
> 
> * Fix a problem where the JVM would restart at
> certain times when using the
>   system time based timer due to an overflow error. 
> This problem was
>   introduced in 3.1.0.  Due to a separate bug in
> 3.1.0, the Wrapper would
>   shutdown rather than simply restarting the JVM as
> was happening in 3.1.1.
>   The last restart happened on Aug 21, 2004.  It
> will next occur Oct 10, 
> 2004
>   and repeat at regular intervals.  There are no
> problems when using the new
>   Tick based timer.  Bug #1014405.
> 
>     Also, what is the reason that you are setting
> your log level to ERROR?
> The way I designed the Wrapper, that is in general a
> bad idea as it 
> hides all
> of the output that you will need to recover from a
> problem.   If the problem
> is that your Java application is sending too much
> output to its console, 
> then
> I would suggest using a Java side logging system
> like log4j or LogKit to
> reduce or redirect that output.
>     Depending on your answer here, I may be able to
> come up with a change
> to the Wrapper to resolve whatever problem you had
> been having as well.
> 
> Cheers,
> Leif
> 
> jawad bokhari wrote:
> 
> >I recently noticed an amazingly strange problem in
> >using wrapper for my applicaiton.
> >
> >Application stopped working at 5 different servers
> at
> >sametime. 
> >I couldnt' see any wrapper.log having my log level
> set
> >to ERROR. I dont' see any error in my application's
> >log.
> >
> >But it suddenly happenend so. 
> >I wonder if there's some ping like application that
> >keep monitoring the service and stops the service
> >externally.
> >
> >Anything I should investigate in wrapper
> >configurations?
> >
> >
> >Regards,
> >
> >Jawad
> >  
> >
> 
> 
> 
>
-------------------------------------------------------
> SF.Net email is sponsored by: Tell us your software
> development plans!
> Take this survey and enter to win a one-year sub to
> SourceForge.net
> Plus IDC's 2005 look-ahead and a copy of this survey
> Click here to start! 
> http://www.idcswdc.com/cgi-bin/survey?id=105hix
> _______________________________________________
> Wrapper-user mailing list
> Wra...@li...
>
https://lists.sourceforge.net/lists/listinfo/wrapper-user
> 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [Wrapper-user] Troubles with use of "tail -1" in "sh.script.in" with GNU coreutils v5.2.1

[Leif M. <le...@ta...>]

Andy,
    That particular line of code had been written by a now inactive 
member of the
Wrapper team.  On linux at least both:
                pid=`$PSEXE -p $pid | grep $pid | grep -v grep | awk 
'{print $1}' | tail -1`
and
                pid=`$PSEXE -p $pid | grep $pid | grep -v grep | awk 
'{print $1}'`
seem to work the same whether the pid exists or not, and whether the 
process has
and child processes or not. Do you see any problems that I am unaware of 
with
this change?  I'll play with it on a Solaris system later today as well.

Cheers,
Leif

Andy Barnett wrote:

> I have encountered troubles using the 
> "{WRAPPER_HOME}/src/bin/sh.script.in" file because of the getpid() and 
> testpid() functions use of "tail -1".
>
> On my SUSE Linux system, the command "tail -1" returns the following:
>
>> tail: `-1' option is obsolete; use `-n 1'
>> Try `tail --help' for more information.
>
>
> And that output causes the "sh.script.in" file to get confused and not 
> find the PID of the running wrapper process.
>
> What that message means is that my SUSE Linux system is preferring the 
> command "tail -n 1".  Unfortunately, this doesn't work on my Solaris 
> system because is doesn't understand the "-n" argument.
>
> On both my SUSE Linux and Solaris systems the "tail" command is part 
> of the GNU coreutils package [1], but I've got version 5.2.1 installed 
> on the SUSE Linux system and 5.0 installed on the Solaris system.
>
> So anyone upgrading their *nix system's GNU coreutils package might 
> encounter this issue.
>
> In comparison, my Mac OS X system happily accepts both "tail -1" and 
> "tail -n 1" as its "tail" command is part of the GNU textutils 
> package, v2.1 (installed via Fink [2])
>
> I don't know that this counts as a bug in the script but I wanted to 
> get it out there so others were aware of it.
>
> My recommendation is that the "sh.script.in" file be updated to use 
> "tail -n 1" and further recommend anyone running the wrapper under a 
> *nix system to be sure and install/upgrade to GNU's coreutils package, 
> v5.2.1 or later.
>
> Cheers,
> ~Andy

Re: [Wrapper-user] Application Stopped Unexpected and amazingly

[Leif M. <le...@ta...>]

Jawad,
    What version are you using? The only thing that I am aware of that could
have possibly caused that is a timing bug in bug in 3.1.0 and 3.1.1.  It was
fixed in 3.1.2.  From the release notes:

* Fix a problem where the JVM would restart at certain times when using the
  system time based timer due to an overflow error.  This problem was
  introduced in 3.1.0.  Due to a separate bug in 3.1.0, the Wrapper would
  shutdown rather than simply restarting the JVM as was happening in 3.1.1.
  The last restart happened on Aug 21, 2004.  It will next occur Oct 10, 
2004
  and repeat at regular intervals.  There are no problems when using the new
  Tick based timer.  Bug #1014405.

    Also, what is the reason that you are setting your log level to ERROR?
The way I designed the Wrapper, that is in general a bad idea as it 
hides all
of the output that you will need to recover from a problem.   If the problem
is that your Java application is sending too much output to its console, 
then
I would suggest using a Java side logging system like log4j or LogKit to
reduce or redirect that output.
    Depending on your answer here, I may be able to come up with a change
to the Wrapper to resolve whatever problem you had been having as well.

Cheers,
Leif

jawad bokhari wrote:

>I recently noticed an amazingly strange problem in
>using wrapper for my applicaiton.
>
>Application stopped working at 5 different servers at
>sametime. 
>I couldnt' see any wrapper.log having my log level set
>to ERROR. I dont' see any error in my application's
>log.
>
>But it suddenly happenend so. 
>I wonder if there's some ping like application that
>keep monitoring the service and stops the service
>externally.
>
>Anything I should investigate in wrapper
>configurations?
>
>
>Regards,
>
>Jawad
>  
>

[Wrapper-user] Application Stopped Unexpected and amazingly

[jawad b. <bok...@ya...>]

I recently noticed an amazingly strange problem in
using wrapper for my applicaiton.

Application stopped working at 5 different servers at
sametime. 
I couldnt' see any wrapper.log having my log level set
to ERROR. I dont' see any error in my application's
log.

But it suddenly happenend so. 
I wonder if there's some ping like application that
keep monitoring the service and stops the service
externally.

Anything I should investigate in wrapper
configurations?


Regards,

Jawad

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[Wrapper-user] Troubles with use of "tail -1" in "sh.script.in" with GNU coreutils v5.2.1

[Andy B. <aba...@ca...>]

I have encountered troubles using the 
"{WRAPPER_HOME}/src/bin/sh.script.in" file because of the getpid() and 
testpid() functions use of "tail -1".

On my SUSE Linux system, the command "tail -1" returns the following:

> tail: `-1' option is obsolete; use `-n 1'
> Try `tail --help' for more information.

And that output causes the "sh.script.in" file to get confused and not 
find the PID of the running wrapper process.

What that message means is that my SUSE Linux system is preferring the 
command "tail -n 1".  Unfortunately, this doesn't work on my Solaris 
system because is doesn't understand the "-n" argument.

On both my SUSE Linux and Solaris systems the "tail" command is part of 
the GNU coreutils package [1], but I've got version 5.2.1 installed on 
the SUSE Linux system and 5.0 installed on the Solaris system.

So anyone upgrading their *nix system's GNU coreutils package might 
encounter this issue.

In comparison, my Mac OS X system happily accepts both "tail -1" and 
"tail -n 1" as its "tail" command is part of the GNU textutils package, 
v2.1 (installed via Fink [2])

I don't know that this counts as a bug in the script but I wanted to 
get it out there so others were aware of it.

My recommendation is that the "sh.script.in" file be updated to use 
"tail -n 1" and further recommend anyone running the wrapper under a 
*nix system to be sure and install/upgrade to GNU's coreutils package, 
v5.2.1 or later.

Cheers,
~Andy

[1] http://www.gnu.org/software/coreutils/
[2] http://fink.sourceforge.net/

Re: [Wrapper-user] Wrapper under Solaris 64 bits

[Leif M. <le...@ta...>]

Alex,

Alexandre Morin wrote:

> Hi, I'm using the service wrapper under Solaris 9 (64 bits).
>
> First, I get this warning which, I guess is because the .so is 32bits (?)
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | WARNING - Unable to load the
> Wrapper's native library 'libwrapper.so'.
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | The file is located on the path
> at the following location but
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | could not be loaded:
>
> INFO | jvm 1 | 2005/03/23 17:38:03 |
> /opt/InfoVista/VistaMart/wrapper/libwrapper.so
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | Please verify that the file is
> readable by the current user
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | and that the file has not been
> corrupted in any way.
>
> INFO | jvm 1 | 2005/03/23 17:38:03 | System signals will not be
> handled correctly.
>
As you guessed, this is being caused because you are running a 64-bit JVM
with a 32-bit JNI library. You will need to compile a 64-bit version of the
Wrapper and use that. The wrapper binary should be fine as 32-bit as it is
a separate process, but the library needs to match the JVM.

If you have the experience, I would appreciate it if you could try modifying
the Makefile.solaris file that ships with the source and coming up with a
Makefile.solaris64 which I can include in future releases of the Wrapper.

> But the problem is that it seem that it is not possible to use more
> than 4Gb of memory (values above get truncated down to 4Gb). Is there
> a workaround to go beyond the 4Gb limit?
>
The Wrapper is actually doing some error checking on the properties
right now.
If you use the wrapper.java.maxmemory property to set the memory then it is
being forced to a range of 1..4096 MB. I need to be checking whether the
version of the Wrapper is 32 or 64 bits. If it is the later, this can
obviously
be much larger.

For now, if you are using 3.1.0 or later, you can set the memory settings
as you would normally by omitting the wrapper.java.maxmemory property
and then using the following instead.
wrapper.java.additional.1=-Xmx6000m

Alternatively, if you are building from source (as above) Then you may just
want to increase the maximum range. Search for 4096 in the wrapper.c
source file. There are 2 places where the upper limit is being forced to
4096.

I don't have a system that I an try this on, so post back with the results.

Cheers,
Leif

[Wrapper-user] Wrapper under Solaris 64 bits

[Alexandre M. <am...@in...>]

Hi, I'm using the service wrapper under Solaris 9 (64 bits).

 

First, I get this warning which, I guess is because the .so is 32bits (?)

 

INFO   | jvm 1    | 2005/03/23 17:38:03 | WARNING - Unable to load the
Wrapper's native library 'libwrapper.so'.

INFO   | jvm 1    | 2005/03/23 17:38:03 |           The file is located on
the path at the following location but

INFO   | jvm 1    | 2005/03/23 17:38:03 |           could not be loaded:

INFO   | jvm 1    | 2005/03/23 17:38:03 |
/opt/InfoVista/VistaMart/wrapper/libwrapper.so

INFO   | jvm 1    | 2005/03/23 17:38:03 |           Please verify that the
file is readable by the current user

INFO   | jvm 1    | 2005/03/23 17:38:03 |           and that the file has
not been corrupted in any way.

INFO   | jvm 1    | 2005/03/23 17:38:03 |           System signals will not
be handled correctly.

 

But the problem is that it seem that it is not possible to use more than 4Gb
of memory (values above get truncated down to 4Gb). Is there a workaround to
go beyond the 4Gb limit?

 

-Alex

 

[Wrapper-user] RE: 2 instances of wrapper on the same machine (on sun and linux)

[EXT-Smith, E. M <eri...@bo...>]

I can confirm that this configuration (multiple services using same
wrapper & libs with different conf files) works very well on Linux,
Solaris, and HP-UX as well as Windows.  I have had between 3 and 5
services running on each of these platforms at the same time.  If you
use the remote interface into the service to shut it down, then you need
different ports, but if you use a control panes (like the service
manager in windows or the xservice manager in Linux, then your wrapper
should trap the kill signal properly and terminate as needed.

Eric M. Smith

Re: [Wrapper-user] Shutdown Hook Security Queston

[Leif M. <le...@ta...>]

Emory,
    Going back and rereading that, it isn't very clear is it.   I had 
written that a long
time ago and to be honest am not sure what I was thinking at the time :-/
    I think the "security" issue would be cases where user code in a 
Tomcat container
or something were to call System.exit.   This would make sure that the 
server came
back up.

    That example would be much better implemented using the 
wrapper.on_exit.<n>
properties however.

    I modified the documentation for the next release to try and clear 
this up.

    Cheers,
    Leif

Emory Guest wrote:

>Hi all,
>
>Under wrapper.disable_shutdown_hook, the description says, "... You may wish to disable the shutdown hook for a number of reasons, including security."
>
>What are the security issues related to the shutdown hook?
>
>Regards,
>
>Em
>  
>

Re: [Wrapper-user] Wrapper in Solaris 9 and 10, Any Problems or Issues?

[Leif M. <le...@ta...>]

Emory,
    As you have run the test code under 9 and 10 I will add those to the 
list of supported
versions.  I don't personally have access to them so was unable to say 
for sure that they
worked.

Cheers,
Leif

Emory Guest wrote:

>Hi all,
>
>The Java Service Wrapper home page lists "Solaris - Sun OS, Solaris 7, 8." under Supported Platforms.
>
>I've run the wrapper test code (and some prototype code of my own) under Solaris 9 and Solaris 10 with no problems. I've searched the mailing list archives for any mention of Solaris 9 and Solaris 10.
>
>Question: Is anyone aware of any issues about using the Java Service Wrapper under Solaris 9 or Solaris 10?
>
>Regards,
>
>Em
>  
>

Re: [Wrapper-user] Can't find dependent libraries?

[Leif M. <le...@ta...>]

Andrew,
    DLLs that are loaded directly as JNI libraries are specified on the 
java.library.path
using the wrapper.java.library.path.n properties.   But if that JNI DLL 
itself requires
other DLLs to be loaded then those other DLLs are located using the 
system PATH.
I am not sure if this is your problem, but if so, try the following:

set.PATH=..%WRAPPER_FILE_SEPARATOR%lib%WRAPPER_PATH_SEPARATOR%%PATH%
wrapper.java.library.path.1=../lib

    The above assumes that all of the DLLs are located in the ../lib 
directory.  The
long WRAPPER_* environment variables are created by the wrapper to make it
possible to make the setting platform independent.  You can also just type:

set.PATH=../lib;%PATH%

Cheers,
Leif

Andrew Dickson wrote:

>I have a Java application which runs fine on 2003 Server (JRE 1.4.2_04). 
>However, when I use the Java Service Wrapper 
>(http://wrapper.tanukisoftware.org/doc/english/introduction.html) to run the 
>application as a service, I get a linking error. The error is definately within 
>my application (nlsxbe.dll), and not the wrapper. The runtime path is exactly 
>the same in both instances (set by wrapper.java.library.path and confirmed by 
>java.library.path), and there is obviously no missing DLL, otherwise the 
>application would not run at all. I do not have this problem on either NT or 
>2000. What should I do to investigate this issue?
>
>
>
>-------------------------------------------------------
>SF.Net email is sponsored by: Tell us your software development plans!
>Take this survey and enter to win a one-year sub to SourceForge.net
>Plus IDC's 2005 look-ahead and a copy of this survey
>Click here to start!  http://www.idcswdc.com/cgi-bin/survey?id=105hix
>_______________________________________________
>Wrapper-user mailing list
>Wra...@li...
>https://lists.sourceforge.net/lists/listinfo/wrapper-user
>
>  
>

[Wrapper-user] Can't find dependent libraries?

[Andrew D. <and...@hs...>]

I have a Java application which runs fine on 2003 Server (JRE 1.4.2_04). 
However, when I use the Java Service Wrapper 
(http://wrapper.tanukisoftware.org/doc/english/introduction.html) to run the 
application as a service, I get a linking error. The error is definately within 
my application (nlsxbe.dll), and not the wrapper. The runtime path is exactly 
the same in both instances (set by wrapper.java.library.path and confirmed by 
java.library.path), and there is obviously no missing DLL, otherwise the 
application would not run at all. I do not have this problem on either NT or 
2000. What should I do to investigate this issue?

[Wrapper-user] Shutdown Hook Security Queston

[Emory G. <emg...@sb...>]

Hi all,

Under wrapper.disable_shutdown_hook, the description says, "... You may wish to disable the shutdown hook for a number of reasons, including security."

What are the security issues related to the shutdown hook?

Regards,

Em

[Wrapper-user] Wrapper in Solaris 9 and 10, Any Problems or Issues?

[Emory G. <emg...@sb...>]

Hi all,

The Java Service Wrapper home page lists "Solaris - Sun OS, Solaris 7, 8." under Supported Platforms.

I've run the wrapper test code (and some prototype code of my own) under Solaris 9 and Solaris 10 with no problems. I've searched the mailing list archives for any mention of Solaris 9 and Solaris 10.

Question: Is anyone aware of any issues about using the Java Service Wrapper under Solaris 9 or Solaris 10?

Regards,

Em

[Fwd: [Wrapper-user] [Fwd: Parameters with spaces and quotes, what a nightmare!]]

[Eduardo B. <e-b...@ad...>]

All,

Please disregard this message. I got my hands on the JAR and fixed the 
root problem (spaces and quotes). Now I'm having a dll problem and am 
trying to fix it following this link:
http://sourceforge.net/mailarchive/message.php?msg_id=11456104

Cheers,

--eb

-------- Original Message --------
Subject: 	[Wrapper-user] [Fwd: Parameters with spaces and quotes, what a 
nightmare!]
Date: 	Tue, 26 Apr 2005 14:33:42 -0500
From: 	Eduardo Basurto <e-b...@ad...>
Reply-To: 	wra...@li...
To: 	wra...@li...



Sorry, forgot to include the escaped version....

Example 4:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2=\"param1=w\"
wrapper.app.parameter.3=\"param2
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4=arg1=x\"
wrapper.app.parameter.4.stripquotes=FALSE
wrapper.app.parameter.5=\"param3
wrapper.app.parameter.5.stripquotes=FALSE
wrapper.app.parameter.6=arg1=y
wrapper.app.parameter.6.stripquotes=FALSE
wrapper.app.parameter.7=arg2=z\"
wrapper.app.parameter.7.stripquotes=FALSE

Cheers,

--eb




-------- Original Message --------
Subject: 	Parameters with spaces and quotes, what a nightmare!
Date: 	Tue, 26 Apr 2005 11:52:40 -0500
From: 	Eduardo Basurto <e-b...@ad...>
To: 	wra...@li...



Hi all,

I am new to the java wrapper and have been reading about ways to configure it , BTW I think is great. But, I have a stubborn jar that requires parameters with spaces and inside double quotes like (from the prompt this works.):
java -classpath %CPATH% com.name.class "param1=w" "param2 agr1=x" "param3 arg1=y arg2=z"

So far, I have tried different ways to pass these parameters in the conf file, but I can't seam to get it working.

Example 1:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2="param1=w"
wrapper.app.parameter.2.stripquotes=FALSE
wrapper.app.parameter.3="param2 arg1=x"
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4="param3
wrapper.app.parameter.4.stripquotes=FALSE
wrapper.app.parameter.5=arg1=y
wrapper.app.parameter.5.stripquotes=FALSE
wrapper.app.parameter.6=arg2=z"
wrapper.app.parameter.6.stripquotes=FALSE

Example 2:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2="param1=w"
wrapper.app.parameter.2.stripquotes=FALSE
wrapper.app.parameter.3="param2 arg1=x"
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4="param3 arg1=y arg2=z"
wrapper.app.parameter.4.stripquotes=FALSE

Example 3:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2=param1=w
wrapper.app.parameter.3="param2
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4=arg1=x"
wrapper.app.parameter.4.stripquotes=FALSE
wrapper.app.parameter.5="param3
wrapper.app.parameter.5.stripquotes=FALSE
wrapper.app.parameter.6=arg1=y
wrapper.app.parameter.6.stripquotes=FALSE
wrapper.app.parameter.7=arg2=z"
wrapper.app.parameter.7.stripquotes=FALSE

The double quotes seam to be required by the class, even:
java -classpath %CPATH% com.name.class param1=w "param2 arg1=x" "param3 arg1=y arg2=z"
will not work, because it requires the double quotes for param1!

I really, really hate spaces and double quotes; but this is my nightmare right now....

Any suggestions?

Thanx in advance.

--Eduardo Basurto

[Wrapper-user] [Fwd: Parameters with spaces and quotes, what a nightmare!]

[Eduardo B. <e-b...@ad...>]

Sorry, forgot to include the escaped version....

Example 4:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2=\"param1=w\"
wrapper.app.parameter.3=\"param2
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4=arg1=x\"
wrapper.app.parameter.4.stripquotes=FALSE
wrapper.app.parameter.5=\"param3
wrapper.app.parameter.5.stripquotes=FALSE
wrapper.app.parameter.6=arg1=y
wrapper.app.parameter.6.stripquotes=FALSE
wrapper.app.parameter.7=arg2=z\"
wrapper.app.parameter.7.stripquotes=FALSE

Cheers,

--eb




-------- Original Message --------
Subject: 	Parameters with spaces and quotes, what a nightmare!
Date: 	Tue, 26 Apr 2005 11:52:40 -0500
From: 	Eduardo Basurto <e-b...@ad...>
To: 	wra...@li...



Hi all,

I am new to the java wrapper and have been reading about ways to configure it , BTW I think is great. But, I have a stubborn jar that requires parameters with spaces and inside double quotes like (from the prompt this works.):
java -classpath %CPATH% com.name.class "param1=w" "param2 agr1=x" "param3 arg1=y arg2=z"

So far, I have tried different ways to pass these parameters in the conf file, but I can't seam to get it working.

Example 1:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2="param1=w"
wrapper.app.parameter.2.stripquotes=FALSE
wrapper.app.parameter.3="param2 arg1=x"
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4="param3
wrapper.app.parameter.4.stripquotes=FALSE
wrapper.app.parameter.5=arg1=y
wrapper.app.parameter.5.stripquotes=FALSE
wrapper.app.parameter.6=arg2=z"
wrapper.app.parameter.6.stripquotes=FALSE

Example 2:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2="param1=w"
wrapper.app.parameter.2.stripquotes=FALSE
wrapper.app.parameter.3="param2 arg1=x"
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4="param3 arg1=y arg2=z"
wrapper.app.parameter.4.stripquotes=FALSE

Example 3:
wrapper.app.parameter.1=com.name.class
wrapper.app.parameter.2=param1=w
wrapper.app.parameter.3="param2
wrapper.app.parameter.3.stripquotes=FALSE
wrapper.app.parameter.4=arg1=x"
wrapper.app.parameter.4.stripquotes=FALSE
wrapper.app.parameter.5="param3
wrapper.app.parameter.5.stripquotes=FALSE
wrapper.app.parameter.6=arg1=y
wrapper.app.parameter.6.stripquotes=FALSE
wrapper.app.parameter.7=arg2=z"
wrapper.app.parameter.7.stripquotes=FALSE

The double quotes seam to be required by the class, even:
java -classpath %CPATH% com.name.class param1=w "param2 arg1=x" "param3 arg1=y arg2=z"
will not work, because it requires the double quotes for param1!

I really, really hate spaces and double quotes; but this is my nightmare right now....

Any suggestions?

Thanx in advance.

--Eduardo Basurto

RE: [Wrapper-user] Configuration File - Space Character Question

[Dick, B. E. <Bri...@FM...>]

Use the short name version of the file. An easy way to get the short
name is to run the following in the directory of the file.

for %i in (YOURFILE) do echo %~si

-----Original Message-----
From: Paul Danckaert [mailto:pa...@le...]=20
Sent: Tuesday, April 26, 2005 9:53 AM
To: wra...@li...
Subject: [Wrapper-user] Configuration File - Space Character Question


Hi,

I've got wrapper working quite well in my testing thus far.. thanks for=20
a nice product!  When I shifted to test my application on Windows, I=20
ran into a problem where wrapper basically blows up when a space=20
character is in a file name/path.  To be fair, I don't know that this=20
is really wrappers fault, but I have a line like this:

wrapper.java.additional.1=3D-Dmyapp.root=3Dd:\Program Files\My =
Application\

For example, and it will simply error and not pull the whole element=20
together.  I've tried variations by putting quotes around the file=20
path, the whole definition, and double quoting (which I understand is a=20
windows batch file escape to get a quote) but nothing seems to work.

Does anybody have a pointer on how to solve this problem? I haven't=20
tested this on unix, but I don't normally have spaces in my directories=20
there.

thanks,
paul



-------------------------------------------------------
SF.Net email is sponsored by: Tell us your software development plans!
Take this survey and enter to win a one-year sub to SourceForge.net
Plus IDC's 2005 look-ahead and a copy of this survey
Click here to start!  http://www.idcswdc.com/cgi-bin/survey?id=3D105hix
_______________________________________________
Wrapper-user mailing list
Wra...@li...
https://lists.sourceforge.net/lists/listinfo/wrapper-user