Hello guys,

Good news!

The following installers are fixed:
   - Mac OS X
   - Debian
   - Windows
   - Binary
   - Archive

I now need to test the two remaining ones:
   - Solaris
   - Fedora

Regards,
Pierre-Arnaud


On 9 mars 2010, at 15:42, Pierre-Arnaud Marcelot wrote:

> Hi again guys,
> 
> Actually it seems the Tanuki issue is just a warning a we could let it like 
> that (more testing needed).
> However, we won't be able to move to a newer version of Tanuki because its 
> licensing have changed and it's now using GPLv2.
> The version we're currently using is the last one that shipped with a 
> MIT-like license (compatible with our Apache License 2).
> It means that we won't be able to support natively 64 bit JVM on Mac OS X 
> (but it might work well with the 32-bit version of the wrapper)
> 
> So, the Tanuki wrapper is not cause of Apache DS failing to load correctly...
> Looking back at the wrapper logs, the last lines are the ones that matter:
>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>> STATUS | wrapper  | 2010/03/09 11:24:38 | --> Wrapper Started as Console
>> [...]
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | Error in WrapperListener.start 
>> callback.  java.lang.NullPointerException
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | java.lang.NullPointerException
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> java.io.File.<init>(File.java:222)
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> org.apache.directory.daemon.Bootstrapper.setInstallationLayout(Bootstrapper.java:83)
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> org.apache.directory.daemon.TanukiBootstrapper.start(TanukiBootstrapper.java:52)
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> org.tanukisoftware.wrapper.WrapperManager$12.run(WrapperManager.java:2788)
>> STATUS | wrapper  | 2010/03/09 11:24:41 | <-- Wrapper Stopped
>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> 
> 
> There's a NPE in the Bootstrapper class at line 83:
>>       File instanceHome = new File( System.getProperty( "INSTANCE_HOME" ) );
> 
> 
> It looks like the INSTANCE_HOME system property is null.
> This system property as been added and the code of the Bootstrapper class has 
> been modified after the last release (1.5.5) in order to support the multiple 
> instances Installation Layout.
> Now, we have problem with this because the same name is already used in 
> various installers to represent the instances home directory (where instances 
> reside).
> This property should be set by the wrapper (even if it's in a wrong context 
> and with a wrong value) when launching the server with the following command 
> line call:
>> "/usr/local/apacheds-1.5.6-SNAPSHOT/bin/apacheds" 
>> "/usr/local/apacheds-1.5.6-SNAPSHOT/conf/apacheds.conf" 
>> set.INSTANCE_HOME=/usr/local/apacheds-1.5.6-SNAPSHOT/instances 
>> set.INSTANCE=default wrapper.syslog.ident=apacheds 
>> wrapper.pidfile="/usr/local/apacheds-1.5.6-SNAPSHOT/instances/default/default.pid"
>>  wrapper.daemonize=TRUE
> 
> 
> However, what we get in the Bootstrapper for this property is null and thus 
> the NPE.
> 
> I'm currently trying to understand why this value could be null...
> 
> 
> I'm afraid that most of our installers are broken due to this issue. 
> At least those using the wrapper.
> Other installers directly launching the server from java like the archive 
> installers seem fine...
> 
> Until this issue is fixed, Apache DS can't be released... :(
> 
> Regards,
> Pierre-Arnaud
> 
> On 9 mars 2010, at 11:41, Pierre-Arnaud Marcelot wrote:
> 
>> Hi Guys,
>> 
>> I'm currently reviewing the installers before the release.
>> 
>> I just found that our Mac OS X installer is broken on Snow Leopard systems.
>> On these systems, the JVM is 64-bit while the wrapper binary we use is still 
>> a 32-bit one.
>> 
>> Launching the server ends with an error:
>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>> STATUS | wrapper  | 2010/03/09 11:24:38 | --> Wrapper Started as Console
>> STATUS | wrapper  | 2010/03/09 11:24:39 | Launching a JVM...
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | Wrapper (Version 3.2.3) 
>> http://wrapper.tanukisoftware.org
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |   Copyright 1999-2006 Tanuki 
>> Software, Inc.  All Rights Reserved.
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | 
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | 
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | WARNING - Unable to load the 
>> Wrapper's native library 'libwrapper.jnilib'.
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           The file is located on 
>> the path at the following location but
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           could not be loaded:
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |             
>> /usr/local/apacheds-1.5.6-SNAPSHOT/bin/../lib/libwrapper.jnilib
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           Please verify that the 
>> file is readable by the current user
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           and that the file has 
>> not been corrupted in any way.
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           One common cause of this 
>> problem is running a 32-bit version
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           of the Wrapper with a 
>> 64-bit version of Java, or vica versa.
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           This is a 64-bit JVM.
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           Reported cause:
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |             
>> /usr/local/apacheds-1.5.6-SNAPSHOT/lib/libwrapper.jnilib:  no suitable image 
>> found.  Did find:  /usr/local/apacheds-1.5.6-SNAPSHOT/lib/libwrapper.jnilib: 
>> no matching architecture in universal wrapper
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |           System signals will not 
>> be handled correctly.
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | 
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | Error in WrapperListener.start 
>> callback.  java.lang.NullPointerException
>> INFO   | jvm 1    | 2010/03/09 11:24:40 | java.lang.NullPointerException
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> java.io.File.<init>(File.java:222)
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> org.apache.directory.daemon.Bootstrapper.setInstallationLayout(Bootstrapper.java:83)
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> org.apache.directory.daemon.TanukiBootstrapper.start(TanukiBootstrapper.java:52)
>> INFO   | jvm 1    | 2010/03/09 11:24:40 |    at 
>> org.tanukisoftware.wrapper.WrapperManager$12.run(WrapperManager.java:2788)
>> STATUS | wrapper  | 2010/03/09 11:24:41 | <-- Wrapper Stopped
>> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>> 
>> This will require that we provide two types of installer for Mac OS X:
>>  - one for 32-bit JVMs
>>  - one for 64-bit JVMs
>> 
>> And more importantly this will force us to move to a newer version of the 
>> Tanuki Wrapper because a 64-bit version of the wrapper for Mac OS X does not 
>> exist for the 3.2.3 version we're using.
>> Now, I remember there has been some problems with Tanuki's licensing when 
>> they changed it a couple of years ago I think...
>> Does anyone know if their Community editor is compatible with the Apache 
>> License ?
>> 
>> Thanks,
>> Pierre-Arnaud
>> 
>> 
>> On 8 mars 2010, at 09:32, Alex Karasulu wrote:
>> 
>>> Yeap this search performance issue will make this release useless.
>>> 
>>> On Sun, Mar 7, 2010 at 9:37 PM, Emmanuel Lecharny <elecha...@gmail.com> 
>>> wrote:
>>>> On 3/7/10 8:21 PM, Stefan Seelmann wrote:
>>>>> 
>>>>> Emmanuel Lecharny schrieb:
>>>>>> 
>>>>>> On 3/7/10 12:33 PM, Stefan Seelmann wrote:
>>>>>>> 
>>>>>>> I built the installers and tested the zip archive. When starting the
>>>>>>> server I get lot of warn logs [1]:
>>>>>>> - "setting the schema manager for partition...". I think those are only
>>>>>>> info messages so should be logged in info log level.
>>>>>>> - "Cannot normalize the value". Are that serious issues? If not we
>>>>>>> should change the logging. I'm sure some users will complain and ask on 
>>>>>>> the
>>>>>>> mailing list what that output means.
>>>>>> 
>>>>>> I will investigate thse two issues. At first sight, it seems tat the
>>>>>> first one is clearly DEBUG or INFO logs wrongly using WARN level.
>>>>>> 
>>>>> I will hold the vote once again. Beside the mentioned issues there is also
>>>>> a serious search performance problem.
>>>> 
>>>> I agree. I'm also fixing some logs issues we have.
>>>> 
>>>> Sorry, Kiran, not your fault :/
>>>> 
>>>> --
>>>> Regards,
>>>> Cordialement,
>>>> Emmanuel Lécharny
>>>> www.nextury.com
>>>> 
>>>> 
>>>> 
>>> 
>>> 
>>> 
>>> -- 
>>> Alex Karasulu
>>> My Blog :: http://www.jroller.com/akarasulu/
>>> Apache Directory Server :: http://directory.apache.org
>>> Apache MINA :: http://mina.apache.org
>> 
> 

Reply via email to