RE: [JBoss-dev] Release 2.4.5 ??

2002-03-25 Thread Coetmeur, Alain

by the way,
 could someone integrate my patch
https://sourceforge.net/tracker/index.php?func=detailaid=512854group_id=22
866atid=376687

It seems to work well (the main problem is explaing what/how you can do with
it), 
but despite the fact I'm allowed to write,
I cannot use CVS because of our firewall...

moreover I've tested all with jboss2.4.4 (+tomcat 4.0.1 .2 .3 ) but not in
jboss3.x




 -Message d'origine-
 De: Thieme, Winfried [mailto:[EMAIL PROTECTED]]
 Date: lundi 25 mars 2002 16:39
 À: '[EMAIL PROTECTED]'
 Objet: [JBoss-dev] Release 2.4.5 ??
 
 
 Hello!
 
 Can anyone tell me when JBoss 2.4.5 will be released? Days, 
 weeks, months ??
 I'm waiting for a bug fix of bug #504895 ...
 
 Thanx
 
 Winfried
 
 
 
 
 ___
 Jboss-development mailing list
 [EMAIL PROTECTED]
 https://lists.sourceforge.net/lists/listinfo/jboss-development
 

___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



RE: [JBoss-dev] Question for win32 batch file gurus

2002-02-18 Thread Coetmeur, Alain



 -Message d'origine-
 De: Jason Dillon [mailto:[EMAIL PROTECTED]]
 Date: samedi 16 février 2002 02:30
 À: Jboss-Development@Lists. Sourceforge. Net
 Objet: [JBoss-dev] Question for win32 batch file gurus
 
 
 Is there any way to detect the directory in which a .bat file 
 lives when
 it is running?

%~dp0   or equivalent %~d0%~p0
(translation in human: disk and path of the 0 variable ... you can get also
the name and option)

look in for/? for detail  (why in for/?, because microsoft)

look in set/? call/? for some more...

not that it works only with CMD.EXE NT shell with extension activated (tha
default)..
probably also on 2000, BUT not on 95/98...
this is why it is often not used for most OpenSource project scripts...

for XP ? dunno

classic starting for a runserver.bat :

@echo off
setlocal
set cmd_dir=%~dp0
for /f tokens=* %%i in ('cd') do set current_dir=%%i
title foo bar server

 ...
goto end
:error
echo problem with
pause
:end
echo Bye...
endlocal

 
 Like in sh, I can `dirname $0` to find out where the script lives.
 
 Is this possible?  If so can some one show me the syntax?
 
 --jason
 
 
 
 
 ___
 Jboss-development mailing list
 [EMAIL PROTECTED]
 https://lists.sourceforge.net/lists/listinfo/jboss-development
 

___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



RE: [JBoss-dev] catalina Connector customizing... bug/patch/ideas (was: org/jboss/web/catalina ajp13connector support)

2002-01-25 Thread Coetmeur, Alain

 -Message d'origine-
 De: Luke Taylor [mailto:[EMAIL PROTECTED]]
 Coetmeur, Alain wrote:
   by the way your ideas seems good, and moreover
   to help people whou want to explicitly configure
   each of their catalina connector
   (like the people tryin to have both http and https),
   the type none could be added
   with the only changes at
   if( type.equals(http) == false  type.equals(warp)
  == false 
   type.equals(ajp)==false  type.equals(none) )
 
 
 what about just using
 
 if(type.equals(none))
 
 :)
 #8))
right, in fact 
I have done it yesterday, and this works...

I have experimented few changes:

1- the none kind of connector

it differs from my proposition here in the fact
that I create the catalina engine even for the
none (same code as HTTP connector
but without the connector)

2- allowing Engine, Service Host Server elements in Config attribute

I've found no code in the confighandle to
support Engine elements in the Config attribute.

We always need an Engine, but it may be cool
to allow the user to configure it more precisely.
but if the user is allowed to configure the engine,
the none kind of connector have not to create one.

maybe will I try that soon.
Host elements, may also be interesting to support.
SERVER and SERVICE parameters may too...

3- added MBEan parameters for HTTO-HTTPS redirectPort and for reverse Proxy

I've also patched the catalina MBean so that
it now support
redirectPort, proxyName and proxyServer parameters

but in fact with the None connector type, and
the Config attribute these change are not needed...
all one have to do is to configure
all the connectors he wants in the Config attribute.


by the way I've found 2 problems, solved one,
and turned around the second

4- DOM to SAX conversion 
 (catalina Mbean dependent on crimson implementation)

the Config Handler class used by catalina MBean
have to convert the XML Element from DOM
to Sax, and do that through the toString()
method of DOM Elements that, in crimson parser, serialize
the Element as XML text. then the config handler parse it with Sax.

however with xerces this does not work 
(toString is not in the standard, 
and I need xerces  for cocoon)

my solution have been to implement a DOM to SAX conversion function that
just navigate in the tree and call the SAX handler methods as needed... this
works well...

5- Multiple XML Top Elements in MBean XML Attribute
   are not  supported and the problem is not signaled

the second problem is that only ONE DOM Element
is read by the Attribute  in
the MBean property is a DOM Element...
moreover this is not checked and thus 
the following siblings are silently ignored

by example you can try to configure 2 Connector elements
with catalina MBean, but only the first will be
read !!!

the turnaround is just to create a
SERVICE element that embeds
many connectors...

In fact I did it myself
adding the SERVER and SERVICE 
(which are BTW needed for the SERVER/SERVICE/Connector/Listener element)

eg this config seems to work,
and could serve a a model 
(be carefull, I've not tested all, especially the http-https redirectPort
and the proxy) :

note that the redirectport attribute work only with my own version
of the catalina Mbean
(I'll send it here soon as a proposed patch)

  !-- Embedded Tomcat 4.x - JBossSX SecurityDomain (JaasSecurityDomain) --
  mbean code=org.jboss.security.plugins.JaasSecurityDomain 
 name=Security:name=JaasSecurityDomain,domain=tomcat
constructor
  arg type=java.lang.String value=tomcat /
/constructor
attribute name=KeyStoreURL../conf/tomcat.jks/attribute
attribute name=KeyStorePasschangeit/attribute
  /mbean


  mbean code=org.jboss.web.catalina.EmbeddedCatalinaServiceSX
name=DefaultDomain:service=EmbeddedTomcat 
attribute name=Port80/attribute
attribute name=ConnectorTypehttp/attribute
attribute name=RedirectPort443/attribute
attribute name=Config
 Server 
  Service 
Connector
className=org.apache.catalina.connector.http.HttpConnector
   port=8080 minProcessors=5 maxProcessors=75
   enableLookups=true redirectPort=8443
   acceptCount=10 debug=0 connectionTimeout=6/
Connector 
  className=org.apache.catalina.connector.http.HttpConnector 
  port=8443 minProcessors=5 maxProcessor=75 acceptCount=10
  enableLookups=true scheme=https secure=true 
Factory
className=org.jboss.web.catalina.security.SSLServerSocketFactory
securityDomainName=java:/jaas/tomcat clientAuth=false
protocol=TLS /
/Connector
Connector 
  className=org.apache.catalina.connector.http.HttpConnector 
  port=443 minProcessors=5 maxProcessor=75 acceptCount=10
  enableLookups=true scheme=https secure=true 
Factory
className=org.jboss.web.catalina.security.SSLServerSocketFactory
securityDomainName=java:/jaas/tomcat clientAuth=false
protocol=TLS /
/Connector

RE: [JBoss-dev] org/jboss/web/catalina ajp13connector support

2002-01-24 Thread Coetmeur, Alain


-Message d'origine-

I was wondering if there were a particular reason why 
EmbeddedCatalinaServiceSX was written to support http 
and warp connectors, but not Ajp13.  

I've tried to insert your patch in my environement 
beside other changes of my own, but I can't find AJP packages
in tomcat4.0.1/catalina sources ...

what could be the reasons... is ajp support dropped now in tomcat4.0.1 ?
stange ? maybe is it hidden somewhere or separate ?

I added an initAjpConnector method which takes the default 
Embedded engine and then substitutes 'Ajp' for 'Warp' for 
the connector-specific code, and this seems to work 


  Also, Warp doesn't seem to automatically deliver static content to Apache
and dynamic content 
 to catalina, although that is a promised feature, and I can't direct 
 it to specifically serve all *.jsp files. 


what are the symptoms ?
normally Apache+mod_webapp  should read the static content in the .WAR and 
not in static files, unlike ajp... (I may have wrongly understood the
intended functionalities)


Should I be afraid of using Ajp?




___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



RE: [JBoss-dev] More Scheduler Stuff

2002-01-21 Thread Coetmeur, Alain



 -Message d'origine-
 De: Andreas Schaefer [mailto:[EMAIL PROTECTED]]
 I will add now the feature to the Scheduler to
 call a MBean instead of a Schedulable class.
 This would need two new attributes:
 - setSchedulableMBean: JMX MBean name
 - setSchedulableMBeanMethod: method name and what attributes 
 are to pass
which can be:
- 'Notification' the notification from the Timer
- 'Date' date of the call
- 'Repetitions' number of the remaining repetitions
- 'SchedulerName' MBean name of the Scheduler called
- any Datatype like 'java.lang.String' where a NULL will
   be passed to (which will not work with basic data types)
 
 Any other suggestions or remarks ?

this seems really cool. The Scheduler is an important
tool in an enterprise app... crontab often does the job, but
running inside jboss is an advantage

One need I had was to call some of my own application
EJB from the scheduler. (this would make the application
less dependent on jboss, since an external crontab may call
an EJB method itself)

with your proposed architecture, one would only need
to design a MBean that respect the schdulable mbean
contract, and then should call a  wellknown EJB.

here is a discussion about HOW in my opinion
an EJB should be called (what's your opinion on that?) :

short version:
I propose to support 2 kind of MBean to invoke an EJB periodicaly :
1- an MBean that calls a stateless session bean method.
   - the most simple would be to declare a home/remote interface
globally, that the MBean could call, and the business
session bean could implement. moreover thing could be easily ported
to a n
   -there should exists a parameters in the MBean
   indicating wether to queue, ignore or invoke concurently 
   notifications calls that ovelap.
2- an MBean that send a JMS message to a message bean.
   - message content is not very important, so they will be either empty, 
 constant text (a CDATA ?)
 or just contain the java.util.Date.toString of the notification
date.
   - session factory and topic/destination will be found in JNDI
   - if someone want to change the content, 
either should he derive that MBean, or
use a session bean that does the job...



long version (arguments...):

the most simple would be to call a stateless EJB.
whether to reuse the same session bean or to create one
each time is a performance and reliability 
(what if an EJB is undeployed meanwhile) issue.
another issue is how to behave if one call takes much time,
and if one call overlap another (overlap, queue, ignore)...

calling a stateful EJB could seems useful if one want
to parametrize the call... but deploying a new stateless
session bean with diffrent env-entry does the same and is simpler.
(so I think this is not usefull).

calling an entity bean could be done for two distinc reasons
- to parametrize the notification, with the same opinion as stateful session
  (ei: not very usefull)
- to implement a kind of singleton and control concurent call to
the notification method...  one could implement this control in the
stateless session bean, but as this is a common need, it would be
practical to ease that. however my opinion is that the best place to
do that is in the notification Mbean, or in the scheduler,
that should be configurable to forbide reentry...

calling a message bean through JMS could be a nice feature, because
message bean invokation are naturally queued, and supports 
long operations. parameters is not really a problem since
there are no essential parameters to be transmited...
the mbean could refers to JNDI-stored JMS destinations and factory...
for most application sending an empty message to a dedicated message bean,
or sending a constant text message to a polyvalent message bean
would do the job...
sending the date could be useful in some rare cases, where the initial date
of notification is important, because calls may be queued for a long time...
but anyway if one want to generate complex JMS message 
one won't avoid writing code, either derivating the MBean, or writing a 
stateles session bean that send the JMS...


for session bean the question could be also wether to depend on a static
remote/home interface of to support dynamic calls...
if we wan our application to be easily portable to a classic
crontab of java thread system, the best would be to avoid 
dynamic invocation which is not perfect on some J2EE system...



does it seems reasonable to you ?





___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



RE: [JBoss-dev] Re: scheduler bug?

2002-01-04 Thread Coetmeur, Alain

have you read my message about that?
maybe does it not work (possible
since my code is a variant of schedule)

anyway check recent version, and also check the archive in jboss-user or in
the forum, or in jboss-dev
since people here have already answered me that
it was solved...

anyway, you can correct it the DIY way :
if it is the addnotification dynamic call that 
fails, check the signature of the addNotification,
and modify the dynamic call so that it
match, the number of arguments, the dynamic 
values, and the dynamic types names...

don't forget that basic types have no .class field
(they ar not classes indeed!), but use the .TYPE
fies of the corresponding holding class, eg:
Long.TYPE.getName()
unlike usual
String.class.getName() because String is a class...

also pass the holder type as the dynamic parameters
since the value parameters list is an array of Object,
and basic type are not Object !

depending on the code you have, you should obtain something like
mActualSchedule = ( (Integer) getServer().invoke(
 mTimer,
 addNotification,
 new Object[] {
Schedule,
Scheduler Notification,
null,
new Date( new Date().getTime() + 1000 ),
new Long( mActualSchedulePeriod ),
new Long( mRemainingRepetitions )
 },
 new String[] {
.getClass().getName(),
.getClass().getName(),
Object.class.getName(),
Date.class.getName(),
Long.TYPE.getName(),
Long.TYPE.getName()
 }
  ) ).intValue();


note that, I repeat, this is not the good mailing list...
this is the jboss internal developers list.
I may even be flammed to death for giving an answer here.
jboss-user was the good one but is officialy (not really)
shut-down

forums on jboss.org are the good place, but to be honnest
I don't read it (an old debate about push/pull
that take place
when jboss-user was shut-down)...

anyway read the archives in forums 

 -Message d'origine-
 De: rajeshwarraov [mailto:[EMAIL PROTECTED]]
 Date: vendredi 4 janvier 2002 08:16
 À: [EMAIL PROTECTED]
 Objet: [JBoss-dev] Re: scheduler bug?
 
 
 I am also facing same problem...I need the solution 
 urgently..anybody out there to help me

___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



RE: [JBoss-dev] Re: scheduler bug? probably...

2002-01-04 Thread Coetmeur, Alain

I've read the current (latest) 2.4.4
source of the scheduler 
 this is not the old bug I was talking about, but 
another I had forgotten til now, and corrected by the way...

note that it is still present in CVS, in last version 1.4.4.1 of the 2.4.4
branch
and in last 1.12 version of the 3.0* head branch... 

summary:
in 2 of the 4 dynamic invocation of the timer addnotification method,
the user object parameter is forgotten...

I've attached a version of Scheduler.java that should work
I've not tested myself (it works on my variant of Scheduler )...
if it correct the problem it should be integrated in 2.4.4 and 3.0 CVS
branch

you can now :
- either:
 -make the change according to my hypothesis,
 -or take my code if you are confident
- then either :
  -compile (with jboss.jar, withou the scheduler*.class), rebuild, and
deploy jboss.jar, 
  -compile the class, create a jar patch, and refer it in jboss startup
options
  -change the package name, compile as usual, make a jar,put it in lib/ext,
and use it instead of org.jboss.util.Scheduler

then if ok, could you tell everybody that it works
(if not, tell me, and if all fails, I'll make my public self-criticism here
8 )

if it works this mail could be a patch proposal to jboss-dev community
(IMHO, a good reason to talk about that here)

hope I'm not wrong,
and hope it will help...

--- long version
I've read the code the addNotification refers
to JMX standard MBean interface Timer.

(could you send me the exact error messages...)

the official signatures of addNotification on
javax.management.timer.Timer

http://dyade.inrialpes.fr/aaa/public/java/jmx/jmx_instr_agent_snmp_javadoc/j
mx/index.html


java.lang.Integer addNotification(java.lang.String type, java.lang.String
message, java.lang.Object userData, java.util.Date date) 

java.lang.Integer addNotification(java.lang.String type, java.lang.String
message, java.lang.Object userData, java.util.Date date, long period) 
java.lang.Integer addNotification(java.lang.String type, java.lang.String
message, java.lang.Object userData, java.util.Date date, long period, long
nbOccurences) 

but in the 3rd and 4th dynamic invocation of addNotification there is no
proposed userData object (which is not used here), despite the fact it is
needed in all version of the addNotification

one should propose a null object as 3rd parameter like it is done in the 1st
and 2nd invocation

you should change the callsto
   // Add an initial call
   mActualSchedule = ( (Integer) getServer().invoke(
   mTimer,
   addNotification,
   new Object[]
   {
  Schedule,
  Scheduler Notification,
null, /* user object */
  mStartDate
   },
   new String[]
   {
  String.class.getName(),
  String.class.getName(),
 Object.class.getName(), /* user object
*/
  Date.class.getName(),
   }
   ) ).intValue();

and 
// Add regular schedule
mActualSchedule = ( (Integer) getServer().invoke(
mTimer,
addNotification,
new Object[]
{
   Schedule,
   Scheduler Notification,
null, /* user object */
   new Date( new Date().getTime() + 1000 ),
   new Long( mActualSchedulePeriod ),
   new Long( getRemainingRepetitions() )
},
new String[]
{
   .getClass().getName(),
   .getClass().getName(),
 Object.class.getName(), /* user object
*/
   Date.class.getName(),
   Long.TYPE.getName(),
   Long.TYPE.getName()
}
) ).intValue();


so, get the source of 2.4.4, modify  them accordingly,
or use the version I've attached
and then  rebuild jboss.jar...

 if you don't want to change it
you can :
- get a copy of jboss.jar
- remove Scheduler.class (and Scheduler$SchedulableExample.jar and
Scheduler$Listenr.class) from this jboss.jar
- recompile your version of Scheduler with this jboss.jar 
 and add the compiled classes to  this jboss.jar now, and put it in place of
the former jboss.jar

you can also begin like before with a copy of the jboss.jar, and then 
build a scheduler-patch.jar with scheduler.class, ad put them in a patch
directory
that you indicate with the -p option (I've never tested that)


finally the most easy is to copy the Scheduler.java, correct the bug, change
the
package name (eg: package 

RE: [JBoss-dev] scheduler bug?

2002-01-02 Thread Coetmeur, Alain

first of all, before beeing flamed, I think
that this list is not the good one for
asking support... it is (as far as I've gerd here)
the list for the jboss developers...

then, I think you have found a bug I've already found,
and that should be corrected now in latest release...
it was something about the dynamic type of the basic long parameters
it works with the dynamic type:
Long.TYPE.getName()

I don't remember what it was before but this
code works for me on 2.4.4. check in the CVS repository since
developers here told be it was corrected...
check my previous mail in the archive also.


mActualSchedule = ( (Integer) getServer().invoke(
 mTimer,
 addNotification,
 new Object[] {
Schedule,
Scheduler Notification,
null,
new Date( new Date().getTime() + 1000 ),
new Long( mActualSchedulePeriod ),
new Long( mRemainingRepetitions )
 },
 new String[] {
.getClass().getName(),
.getClass().getName(),
Object.class.getName(),
Date.class.getName(),
Long.TYPE.getName(),
Long.TYPE.getName()
 }
  ) ).intValue();

 -Message d'origine-
 De: Paris Deligiannakis [mailto:[EMAIL PROTECTED]]
 Date: mercredi 2 janvier 2002 16:41
 À: [EMAIL PROTECTED]
 Objet: [JBoss-dev] scheduler bug?
 
 
 I'm not really sure about this, so please do correct me if 
 it's wrong or it has been brought up previously:
 
 in Scheduler.java there is this:
 
 [code]
 mActualSchedule = ( (Integer) getServer().invoke(
   mTimer,
   addNotification,
   new Object[] {
  Schedule,
  Scheduler Notification,
  mStartDate
   },
   new String[] {
  String.class.getName(),
  String.class.getName(),
  Date.class.getName(),
   }
) ).intValue();
 [/code]
 
 However I could not find an addNotification method with only 
 3 arguments (an object is required before the Date arg).
 And when this code runs, it throws this:
 
 [ERROR,Default] javax.management.ReflectionException: The 
 operation with name addNotification could not be found
 
 
 regards,
 paris
 __
 View this jboss-dev thread in the online forums:
 http://jboss.org/forums/thread.jsp?forum=66thread=6463
 
 ___
 Jboss-development mailing list
 [EMAIL PROTECTED]
 https://lists.sourceforge.net/lists/listinfo/jboss-development
 

___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



[JBoss-dev] minor Bug in org.jboss.util.Scheduler v2.5.3

2001-12-10 Thread Coetmeur, Alain

I've found a bug in the
call to removeNotification
in the scheduler MBean

the dynamic parameter type is set as int
instead of class Integer...

there are 3 call like these (bug corrected here) 

getServer().invoke(
   mTimer,
   removeNotification, 
   new Object[] {
  new Integer( mActualSchedule )
   },
   new String[] {
  /* Integer.TYPE.getName() */
  Integer.class.getName()
   }



note that I've found that while building
my own version of this class
that calls a generic session EJB 

I'll soon propose it as a 
org.jboss.contrib.ejbscheduler.EjbScheduler MBean package...

I've also adapted jboss-iiop/jacorb mbean
to have a a working corba nameservice, with persistent store.
I should adapt it to latest jacorb before publishing.

I've done all of that, because 
I'm deploying a demonstation site in my
intranet... it is a portfolio optimization
application, with Jboss/tomcat/jacorb/PostgreSQL/Mico...

congratulation for the job done, thanks for all,
and hope this helps.





--
Alain Coetmeur

___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



RE: [JBoss-dev] minor Bug in org.jboss.util.Scheduler v2.4.3

2001-12-10 Thread Coetmeur, Alain



 -Message d'origine-
 De: Andreas Schaefer [mailto:[EMAIL PROTECTED]]
 Date: lundi 10 décembre 2001 18:50
 À: Coetmeur, Alain; [EMAIL PROTECTED]
 Objet: Re: [JBoss-dev] minor Bug in org.jboss.util.Scheduler v2.5.3
 
 
 Hi

 As far as I see these bugs were fixed at 8/14/2001.
sorry for the noise

  I'll soon propose it as a 
  org.jboss.contrib.ejbscheduler.EjbScheduler MBean package...
 
 Can you explain which this is necessary ?

this came from my own problem : to download data regularily in
 the database I use... All my data are managed by EJB.

si my first solution was to run a java application that
connect to the EJB and do the job...

then, in the hope to package all in Jboss server, I've 
tried to play with the JBOSS Scheduler... 
I did that by adding the ejb-client jar to the MBean path
(with the classpath mbean) and call my loader class which calls the EJB.


however
I've encountered a strange error in the web application ...
Jboss failed on a classcast exception in some EJB methods, in the $Proxy.
in fact some by value classes (classes that are exchanged with content
inside)
were exchanged, but the version in the EJB jar and from the MBean global
context
were different (different classloader I suppose) yet have the same name...
all of that could be corrected by separating the interfaces for my
loader EJB and the one from all the rest, including the values...

However this convinced me to desing something more generic and
less tricky for the user :

I've created a generic EjbSchedulable remote interface (and a home
interface),
with e perform(...) method
that a user EJB may use .
a variant of the Scheduler MBean does a similar job 
(not the Schedulable interface because of RemoteException), 
but calls the session EJB that it find through a JNDI name...


the application that use this EjbScheduler have only
to declare the session EJB and  reference the remote/home interface,
then add a Mbean descriptor with the JNDI name properly set, beside
period and count.

We couls alsp use a message EJB, through JMS
but I've no knowledge in ejb2.0...
maybe useful  to serialize the calls.



___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development



[JBoss-dev] [JBossCMP] W32.Nimda.A@mm worm attack on jbosscmp

2001-09-19 Thread Coetmeur, Alain


I've receives this worm attack (below without the infectious attachments !)
through the jbossCMP mailing list.
It tries to execute a .exe, when you preview the message,
even if you open no attachment at all...

please beware this is a well known worm
http:[EMAIL PROTECTED]
exploiting a Win32/MIME/IE5.x bug
corrected by recent patches
like this one 
http://www.microsoft.com/windows/ie/downloads/critical/q299618/default.asp
(windowsupdate.com propose it )

hope this helps...
victims spread the worm through mail or IIS...

-Message d'origine-
De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] 
Date: mardi 18 septembre 2001 18:20
À: undisclosed-recipients
Objet: [JBossCMP]
usamplesamplesampledesktopdesktopsamplesampledesktopsamplesamplesamplesample
desktopsampledesktopsampledesktopdesktopdesktopsamplesampledesktopdesktopsam
plesampledesktopsampledesktopdesktopdesktopdesktopdesktopdesktopdesktopdeskt
opdesktopsampled




--  Virus Warning Message (on the network)

readme.exe is removed from here because it contains a virus.

-

--  Virus Warning Message (on the network)

Found virus TROJ_NIMDA.A in file readme.exe
The file readme.exe is moved to /etc/iscan/virus/virSIAUdKg3_.

For more information, please contact [EMAIL PROTECTED]

-



[JBoss-dev] [JBossCMP] W32.Nimda.A@mm worm attack on jbosscmp

2001-09-19 Thread Coetmeur, Alain


I've receives this worm attack (below without the infectious attachments !)
through the jbossCMP mailing list.
It tries to execute a .exe, when you preview the message,
even if you open no attachment at all...

please beware this is a well known worm
http:[EMAIL PROTECTED]
exploiting a Win32/MIME/IE5.x bug
corrected by recent patches
like this one 
http://www.microsoft.com/windows/ie/downloads/critical/q299618/default.asp
(windowsupdate.com propose it )

hope this helps...
victims spread the worm through mail or IIS...

-Message d'origine-
De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] 
Date: mardi 18 septembre 2001 18:20
À: undisclosed-recipients
Objet: [JBossCMP]
usamplesamplesampledesktopdesktopsamplesampledesktopsamplesamplesamplesample
desktopsampledesktopsampledesktopdesktopdesktopsamplesampledesktopdesktopsam
plesampledesktopsampledesktopdesktopdesktopdesktopdesktopdesktopdesktopdeskt
opdesktopsampled

___
Jboss-development mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/jboss-development