Re: SQL Server JDBC driver from Microsoft

[Greg Matthews]

www.inetsoftware.de

fast, stable, good support.

we've been using it for over a year now and it works fine.

- Original Message -
From: "SAURUGGER,PETER (A-PaloAlto,ex2)" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, March 16, 2002 5:09 AM
Subject: RE: SQL Server JDBC driver from Microsoft


> weblogic (oops) has had a driver for MS SQL server for a while ... used it
> for a demo once, works fine for basic jdbc, don't know about advanced
> features and performance relative to other drivers
>
> -Original Message-
> From: Robert Virkus [mailto:[EMAIL PROTECTED]]
> Sent: Friday, March 15, 2002 5:55 AM
> To: Orion-Interest
> Subject: Re: SQL Server JDBC driver from Microsoft
>
>
> Hallo Jarrod,
>
> Friday, March 15, 2002, 1:46:36 PM, you wrote:
> JR> has anyone gotten this to work with Orion 1.5.4?
>
> it just works fine over here, though it is said, that it should not be
> used in production-environments. We manage our own connection-pooling
> though and haven't had problems so far... If anyone knows more about
> problems within production-environments, please let us know! :-)
> Compared to the sun.jdbc.odbc.JdbcOdbcDriver we gained about 20%
> performance with the jdbc-driver from ms.
>
> Can anyone suggest a good alternative to the ms-driver? has anyone
> experiences with free-tds?
>
> thanks and greetings
>Robert
>
> --
> Robert Virkus
> scaraboo GmbH
> mobile Entertainment
> Georg-Wulf-Str.4-6
> 28199 Bremen
> Germany
> phone  +49 - (0)421 - 59 67 549
> fax+49 - (0)421 - 59 67 567
> mobile +49 - (0)171 - 35 31 635
> [EMAIL PROTECTED]
> www.scaraboo.de
> wap.scaraboo.de
>
> 
> Aus Rechts- und Sicherheitsgruenden ist die in dieser E-Mail gegebene
> Information nicht rechtsverbindlich. Eine rechtsverbindliche Bestaetigung
> reichen wir Ihnen gerne auf Anforderung in schriftlicher Form nach.
Beachten
> Sie bitte, dass jede Form der unautorisierten Nutzung, Veroeffentlichung,
> Vervielfaeltigung oder Weitergabe des Inhalts dieser E-Mail nicht
gestattet
> ist. Diese Nachricht ist ausschliesslich fuer den bezeichneten Adressaten
> oder dessen Vertreter bestimmt. Sollten Sie nicht der vorgesehene Adressat
> dieser E-Mail oder dessen Vertreter sein, so bitten wir Sie, sich mit dem
> Absender der E-Mail in Verbindung zu setzen.
> 
> For legal and security reasons the information provided in this e-mail is
> not legally binding. Upon request we would be pleased to provide you with
a
> legally binding confirmation in written form. Any form of unauthorised
use,
> publication, reproduction, copying or disclosure of the content of this
> e-mail is not permitted. This message is exclusively for the person
> addressed or their representative. If you are not the intended recipient
of
> this message and its contents, please notify the sender immediately.
>
>

Re: Memory leak with linux jdk1.4 - orion 1.5.4

[Greg Matthews]

Robert,

I've monitored orion using jprobe when orion is started with the -restart
command.

It loaded another copy of the application classes in a new classloader, and
eventually ran out of memory.

I never pursued this as i had higher priority things to work on at the time.
Not sure if the original classloader was eventually cleaned up, or whether
there is some config setting that causes Orion to work differently.

Greg.

- Original Message -
From: "Robert Krüger" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Sunday, March 10, 2002 9:06 PM
Subject: Memory leak with linux jdk1.4 - orion 1.5.4


>
> hi,
>
> I'm experiencing serious problems with a memory leak that causes orion to
> consume more memory with each application restart/redeployment with the
> combination  jdk1.4 (linux) - orion 1.5.4. with jdk 1.3 memory consumption
> remains constant after a while. even calling System.gc() explicitly
doesn't
> help. has anyone else seen such behaviour? based on these observations it
> appears more likely to be a VM bug rather than an orion problem. is this a
> known problem? has anyone tried to narrow it down by using a profiling
tool?
>
> tia
>
> robert
>
>

Re: Local interfaces difference ???

[Greg Matthews]

and can anyone shed some light on why this has even been put in the J2EE
spec.

Visibroker for Java automatically detected if the client was local and then
handed out a client stub that did not do marshalling, i.e. performed an
in-process local java method call.

I would have thought that the whole concept of Local references was just an
optimisation that a vendor provided based on common sense.

Why is it in the spec?

Greg.

- Original Message -
From: "Eddie Post" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Thursday, March 07, 2002 4:54 AM
Subject: Local interfaces difference ???


> Hellu,
>
> Can someone tell me what the performance boost is when you change your
> remote and home interfaces to local interfaces with Orion 1.5.4 ?
>
> I did this but I know that with Bea you could already indicate if your
beans
> and clients were located on the same machine such that it already had some
> kind of local workaround. I was wandering how Orion did and does this ??
>
> Eddie
>
>
> _
> Send and receive Hotmail on your mobile device: http://mobile.msn.com
>
>
>

Re: File Upload from an Application

[Greg Matthews]

you need to use
 
1. type="file" in the form within the 
browser.
 
2. get a ServletInputStream (?) from the request 
object in Orion and parse out the byte stream.
 
do a search for ServletInputStream in google or 
something and you should find some examples.
 
afaik, there is no simple method to call to extract 
an uploaded file from a request but parsing out the ServletInputStream data is 
not that hard.
 
greg.

  - Original Message - 
  From: 
  Michael Maurer 
  To: Orion-Interest 
  Sent: Monday, March 04, 2002 5:38 
PM
  Subject: File Upload from an 
  Application
  
  
  Hi 
   
  I am trying to upload a file from 
  an application to the orion server, but could not find the right solution so 
  far. 
  Does anyone has a nice solution 
  ?
   
  Michael

Re: Caching / Pooling of BMP Entity bean.

[Greg Matthews]

it's probably your code. BMP caching works no problem.

your best bet is to post a code example of one of your entity beans.

- Original Message -
From: "Shah, Ritesh" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Thursday, January 17, 2002 8:02 AM
Subject: Caching / Pooling of BMP Entity bean.


> Hi,
>   I am getting a weird error. I am using oracle oc4j 1.0.2.2.
>I am calling an entity bean  from session bean. The error I get is when
I
> call an entity bean for the first time it gives me proper result then I
> update the data and from next call onward it gives me  the cached value
that
> it loaded for the first time call. It even doesn't give me stored value. I
> try to trace it but I get the same error. From database it is retrieving
> proper value while doing ejbFindbyPrimarykey but doesn't give proper
value.
> It never again call ejbload. That what I found by tracing it.
>
> Any help will be appreciated.
>
> Thanks
> -Ritesh
>
>

Re: Simple question re caching/pooling of BMP Entity Beans

[Greg Matthews]

pooling definitely does work for BMP.

couple of things i found that you might want to check off.

1. implement an isModified method

public boolean isModified() {
  return false;
 }

which orion will check to determine if ejbLoad needs to be called. we're
using the value object pattern to the above works for us. you may want to
make it more complicated depending on your requirements.

2.. return a Collection from your finder methods. returning an Enumeration
seems to cause orion to not pool.

3. if you have defined your own primary key instead of using, say, Integer,
it seems you need to implement equals() and hashCode().

if you can't get it working maybe post the code from one of your bean
classes.

good luck,
greg.


- Original Message -
From: "Geoff Soutter" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Friday, January 11, 2002 8:08 AM
Subject: RE: Simple question re caching/pooling of BMP Entity Beans


> Hi Sergey,
>
> Thanks for the reply. I didn't explain it very clearly did I? What I
> meant was:
>
> "However, it appears that it goes to the database (i.e. calls _ejbLoad_
> in the Entity Bean's implementation) for every call to findByPrimaryKey,
> regardless of whether it has previously read the instance with the
> specified key before."
>
> Have you been able to get pooling of BMP entity beans working?
>
> Geoff
>
> > -Original Message-
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]] On Behalf Of
> > Sergey G. Aslanov
> > Sent: Thursday, 10 January, 2002 6:29 PM
> > To: Orion-Interest
> > Subject: Re: Simple question re caching/pooling of BMP Entity Beans
> >
> >
> > Hi Geoff
> >
> > Loading of beans have these stages:
> > 1. calling ejbFindBy... to get the primKeys of the entities;
> > 2. checking if entities with such keys are already pooled;
> > 3. for every key checking: if not pooled - call ejbLoad, else
> > - get instance from pool;
> >
> > As you see, ejbFindBy... are called always independently of
> > entity is pooled or not.
> >
> > Thursday, January 10, 2002, 2:01:37 AM, you wrote:
> >
> > GS> Hi there,
> >
> > GS> I'm teaching myself how to use Orion and I've got a
> > simple question
> > GS> about caching/pooling of BMP Entity Bean instances.
> >
> > GS> I've created a simple app which lets me read a BMP Entity Bean
> > GS> through a Session Bean. I would expect that when I call
> > GS> findByPrimaryKey on my Entity Bean's home interface, Orion caches
> > GS> the instance it creates so that next time I call findByPrimaryKey
> > GS> for the same key, it just returns the instance it already
> > has rather
> > GS> than creating a new one. However, it appears that it goes to the
> > GS> database (i.e. calls ejbFindByPrimaryKey in the Entity Bean's
> > GS> implementation) for every call to findByPrimaryKey, regardless of
> > GS> whether it has previously read the instance with the
> > specified key
> > GS> before.
> >
> > GS> I checked the list archive and it appears that someone
> > has reported
> > GS> a similar problem a while back but there were no replies
> > GS>
> > (http://www.mail-archive.com/orion-interest@orionserver.com/ms
> g17359
> GS> .htm
> GS> l).
>
> GS> I also checked the documentation and for orion-ejb-jar.xml it states
>
> GS> that exclusive-write-access is true by default, and since I am not
> GS> setting it I presume it's defaulting to true which should allow
> GS> Orion to cache the entity beans.
>
> GS> So, can anyone shed any light on how whether/how I can get Orion to
> GS> cache BMP entity beans?
>
> GS> Cheers
>
> GS> Geoff
>
>
> --
> Sergey G. Aslanov,
> CBOSS Group,
> Web-technologies department
> mailto:[EMAIL PROTECTED]
> tel: +7 095 7555655
>
>
>

Re: RuntimeException in threads

[Greg Matthews]

if by user threads you mean a thread that you're manually starting (i.e. not
a worker thead) then maybe put a try/catch block in the run method?

public void run() {
while(!stopped) {
try {
// do stuff
} catch(Exception e) {
// handle error
}
}
}

...or have i missed the point completely?

- Original Message -
From: "Namor Taror" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, November 28, 2001 7:35 PM
Subject: RuntimeException in threads


> I've been trying to catch a runtime exception thrown by one of my user
> threads, before the orion server gets it. One of the solutions I thought
> would work was to create a ThreadGroup as a parent of the thread in
question
> and have it to catch the exception. It doesn't seem to work. Any ideas?
>
> Roman
>
> _
> Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
>
>
>

Re: HttpSession creation

[Greg Matthews]

hi, see Kesav Kumar's answer 7:49pm to learn 
somerthing.

  - Original Message - 
  From: 
  Lachezar 
  Dobrev 
  To: Orion-Interest 
  Sent: Thursday, November 22, 2001 6:15 
  PM
  Subject: Re: HttpSession creation
  
     Hello.
     Trying to stop a session from 
  being created is not valid.
     Sessions are not user-dependant. 
  Sessions are a handshake between a server and a client. You should have 
  thought another way to do that. Maybe check if the user HAS 
  authenticated.
   
     .getSession() and 
  .getSession(boolean) methods are to get access to the session, not to create 
  the session. Creation is left to the App server.
     What happens is, that when a 
  client contacts your Server the session IS created. That's it.
     Forget the option to disable that. 
  It has no sense.
     If you REALLY need to do that... 
  Use "session.invalidate()" everytime you see, that the user has not logged. 
  That way a session will be recreated every time the client 
  connects.
   
   
     Lachezar.
   
     P.S. Sessions are good to store 
  info in them. But user management should NOT be connected with 
it.
   
   
  
hi,
 
i'm trying to ensure that the only time a 
session is created is after a successful user login.
 
is there any other way apart from 
HttpServletRequest.getSession( ), or HttpServletRequest.getSession( boolean 
create) that a HttpSession can be created?
 
i've trawled through the J2EE API but haven't 
seen any other methods that might create a session -- was just wondering if 
there was some sort of side-effect way, say, if you redirect a request or 
pass a request through a filter chain or something???
 
thanks,
greg.

HttpSession creation

[Greg Matthews]

hi,
 
i'm trying to ensure that the only time a session 
is created is after a successful user login.
 
is there any other way apart from 
HttpServletRequest.getSession( ), or HttpServletRequest.getSession( boolean 
create) that a HttpSession can be created?
 
i've trawled through the J2EE API but haven't seen 
any other methods that might create a session -- was just wondering if there was 
some sort of side-effect way, say, if you redirect a request or pass a request 
through a filter chain or something???
 
thanks,
greg.

Re: Most Urgent: How to find jndi lookup name for Transaction Manager?

[Greg Matthews]

InitialContext ic = new InitialContext();
UserTransaction ut =
(javax.transaction.UserTransaction)ic.lookup("java:comp/UserTransaction");


- Original Message -
From: "Ray Harrison" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, November 21, 2001 3:12 AM
Subject: Re: Most Urgent: How to find jndi lookup name for Transaction
Manager?


> Check out the Orionserver FAQ:
> How do I get a reference to the TransactionManager?
>
>
> import javax.naming.*;
> import javax.transaction.*;
> ...
> TransactionManager manager = (TransactionManager)new
> InitialContext().lookup("java:comp/UserTransaction");
>
>
> Cheers
> Ray
>
>
> --- patrick patrick <[EMAIL PROTECTED]> wrote:
> > Dear all,
> >
> >   if any one know how to find the jndi lookup name
> > for the Transaction Manager, please let me know..
> >
> > Thanks
> > patrick
> >
> > __
> > Do You Yahoo!?
> > Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
> > http://geocities.yahoo.com/ps/info1
> >
>
>
> __
> Do You Yahoo!?
> Yahoo! GeoCities - quick and easy web site hosting, just $8.95/month.
> http://geocities.yahoo.com/ps/info1
>

Fw: what the? ejbStore gets called last???

[Greg Matthews]

 
further to this last message, there's information 
on
http://developer.java.sun.com/developer/technicalArticles/ebeans/sevenrules/
about the same issue, i.e. ejbStore being deferred. 
it seems to be rule number 
seven.
 
again, how do you get around the problem where a 
subsequent load of data through a findBy... method is getting the old copy of 
data, because the container has cleverly decided to defer calling 
ejbStore???
 
greg
 
 
- Original Message - 
From: Greg 
Matthews 
To: Orion-Interest 
Sent: Wednesday, October 24, 2001 10:11 AM
Subject: what the? ejbStore gets called last???

 
 

what the? ejbStore gets called last???

[Greg Matthews]

 
dear all,
 
having not used EntityBeans in my last projects, 
but stateless session beans only, i'm hoping someone has a good answer to my 
question.
 
basically, it seems like ejbStore is being deferred 
to the end of the transaction.
 
if i request a series of operations like 
this.
 
1. loadX
2. store (which changes X)
3. loadX
 
then what really happens is
 
1. loadX
2. loadX    <== this gets the old 
copy of X, not the new one since Orion deferred my call to ejbStore
3. store.
 
this post below by someone indicates that jboss 
provided a flag so that the database access occurs in the order that your code 
executes -- as opposed to when the container determines it should 
run.
 
does orion have an equivalent flag? (and yes, i've 
had a look through www.orionserver.com 
and couldn't find it).
 
http://www.mail-archive.com/ejb-container-group2@objectweb.org/msg00011.html
 
thanks,
greg.

Re: Yup, SSL question

[Greg Matthews]

sorry if i'm off track here??, but last time i 
checked, orion couldn't cluster + SSL at the same time.
 
i think it's orion bugzilla defect 525 from 
memory, since i spent some time investigating it. i think elephantwalker is a 
bit of an expert on this particular issue.
 
you can cluster, and you can setup SSL, but just 
not at the same time.
 

  - Original Message - 
  From: 
  Steve 
  Best 
  To: Orion-Interest 
  Sent: Friday, October 19, 2001 4:35 
  AM
  Subject: Re: Yup, SSL question
  
  Nevermind, I have found the keystore and have 
  imported the certificate.  I believe my problems now have to do with our 
  clustering/load balancing configuration.  We are currently using an 
  Alteon Acedirector 3 for our load balancer.  We are also using clustering 
  to maintain sessions.  I think the error I have been 
getting:
   
  "Error listening to SSLServerSocket: No available 
  certificate corresponds to the SSL cipher suites which are 
  enabled."
   
  is because our certificate is for our virtual IP, 
  and not for our local machine, and therefore Orion is unable to load the 
  certificate correctly because it isnt for the local host.  The question 
  is, how do I configure Orion to load a certificate for our virtual site?  
  Here is my configuration, with some entries modified to protect the 
  innocent:
   
         
  load-on-startup="false" />    
          
   
  I have tried setting the web-site host="[ALL]", 
  but I get an error saying I cannot cluster with that configuration.  
  Though I no longer get the SSLServerSocket error.
   
  Thanks for any help,
  Steve
   
  
- Original Message - 
From: 
Steve 
Best 
To: Orion-Interest 
Sent: Wednesday, October 17, 2001 3:43 
PM
Subject: Yup, SSL question

So, I do not have the original keystore file 
generated during the certificate request, but I have the original 
certificate request private key file, as well as the certificate file from 
Thawte.  How do I get them both into a working keystore?  

 
Steve

Re: JSP Compilation Questions and More.

[Greg Matthews]

is this a speed issue?

if so, use jikes. it's so fast that you don't need to precompile JSPs.



- Original Message -
From: "Lachezar Dobrev" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, October 13, 2001 1:50 AM
Subject: JSP Compilation Questions and More.


>Hallo all!
>
>Recently I've been playing around with JSP compilation in Orion. I saw
a
> couple of days ago a question about the out-of-the-server (stand-alone)
> compilation of JSPs. Quite some job :(
>I tried to find the compilator. Playing around I got some things. I
don't
> really remember what was the JSP that helped me, but I guess it is either
> com.evermind._rj (and it's method _bb() ) or com.evermind._ah. Also the
> com.evermind.JspPage has a _bb() method, but it doesn't help a lot.
>
>Question 1: Is there a way to pre-compile the JSPs? I also saw that
> adding a parameter "jsp_precompile=true" to the request parameters will
only
> try to compile the jsp. That however is not what I was thinking about.
>
>Question(BUG) 2: I noticed the following: It depends how the JSPs are
> compiled. If the Web-App is "development=true" it is compiled in one way
> with matching the JSP lines, and if "development=false" the lines are
often
> mapped wrongly. Exa.:
>
> ALABALA.JSP:
>
> 1. 
> 2. 
> 3. Test for test
> 4. 
> 5. 
> 6. 
> 7. 
> 8. <%= new String((String)null)%>
> 9. 
> 10.
>
> Obviously there is a NullPointerException on line 8 (7 if counted from 0).
>
> development = true:
>
> 500 Internal Server Error
> java.lang.NullPointerException
>  at java.lang.String.(String.java:193)
>  at /Alabala.jsp._jspService(/Alabala.jsp.java:24) (JSP page line 7)
>  at com.orionserver[Orion/1.5.2 (build
> 10460)].http.OrionHttpJspPage.service(Unknown Source)
>  at com.evermind[Orion/1.5.2 (build 10460)]._ah._rad(Unknown Source)
>  ..
>
>
> development = false:
>
> 500 Internal Server Error
> java.lang.NullPointerException
>  at java.lang.String.(String.java:193)
>  at /Alabala.jsp._jspService(/Alabala.jsp.java:27) (JSP page line 2)
>  at com.orionserver[Orion/1.5.2 (build
> 10460)].http.OrionHttpJspPage.service(Unknown Source)
>  at com.evermind[Orion/1.5.2 (build 10460)]._ah._rad(Unknown Source)
>  ..
>
>
>Obviously both cases differ. Why?
>
>Lachezar.
>
>
>
>

Re: Session share problem.

[Greg Matthews]

i think there's a share="true" attribute that you 
have to put in the web-site.xml file ??? check out the doco in www.orionserver.com
 

  - Original Message - 
  From: 
  Jishan Li 
  
  To: Orion-Interest 
  Sent: Wednesday, September 05, 2001 7:42 
  PM
  Subject: Session share problem.
  
  Hi,
     I have session share problem between ssl site 
  and non-ssl site. My ssl site name is secure.mysite.com and non-ssl site name 
  is www.mysite.com. when I start my server, 
  and visit www.mysite.com firstly, 
  everything goes well. But when I visit secure.mysite.com firstly after I 
  starting my orion server, and then back to www.mysite.com  every request on www.mysite.com create a new session.  So 
  my user login, shopping cart won't work!!!
     Is there any one can help me?
  Thanks!
   
  Jishan Li.

Re: www.elephantwalker.com

[Greg Matthews]

EW,

did you have to do anything special to get sessions to be replicated across
the cluster?

there have been several posts recently about people having trouble with this
specifically.

do you just have to make sure that all boxes in the cluster have the same
cluster-island id or is something else required?

thanks,
greg/

- Original Message -
From: "The elephantwalker" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, August 25, 2001 4:03 AM
Subject: www.elephantwalker.com


> Dear Group,
>

> 3. The enterprise application is a distributable forum application so that
> if one server goes down, the other server will continue to serve up web
> pages. The users will never know that their server has gone done, since
the
> session are shared across the cluster. This is unlike another well known

Re: Orion Clustering problems..

[Greg Matthews]

kevin,

i've also been able to set up clustering but have not reproduced your error.

i did notice that while i was investigating the various config settings,
there was an entry in rmi.xml that i never used, but i could still cluster,
i.e. load balancer redirected my request to one of the orion boxes, and i
could kill a box, then login again, and my request would be sent to box2.

i never did work out what this setting did.

do you know what this does? does it maybe handle transmission of session
data between boxes?

remember, according to the docs you need boxes that are supposed to share
session info to have the same cluster-island id.

  
  Tag that is defined if the application is to be clustered. Used to set
up a local multicast cluster. A username and password used for the servers
to intercommunicate also needs to be specified.

  host - The multicast host/ip to transmit and receive cluster data on.
The default is 230.0.0.1.
  id - The id (number) of this cluster node to identify itself with in
the cluster. The default is based on local machine IP.
  password - The password configured for cluster access. Needs to match
that of the other nodes in the cluster.
  port - The port to transmit and receive cluster data on. The default
is 9127.
  username - The username configured for cluster access. Needs to match
that of the other nodes in the cluster.



i'd be very interested if you're able to get session clustering/failover
working.

cheers,
greg.

- Original Message -
From: "Duffey, Kevin" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Thursday, August 23, 2001 12:27 PM
Subject: Orion Clustering problems..


> HI all,
>
> Hopefully this will make the list.
>
> I am working on a chapter for a book about scalability. I am also needing
to
> get our site ready for scaling. I am using Orion to test my web-app that
> needs to be a demonstration of a scalable Servlet 2.2 web application. I
> have a couple of questions to ask about session fail-over, and how to get
> Orion to do this.
>
> I followed the cluster doc to the tee, and everything works fine except
that
> the SessionServlet example does NOT keep the count going when I shut down
> the server that the first request to it created a session on. It starts
over
> on the count. So I assume either my config is somehow just not correct, or
> that this is a bug with Orion. Here is what I have in my web-site.xml
file:
>

Re: Finding Default DataSource in My App

[Greg Matthews]

a starting point might be.

import com.evermind.server.administration.*;

javax.naming.Context ctx = new InitialContext();
   ApplicationAdministrator admin =
(ApplicationAdministrator)ctx.lookup("java:comp/Administrator");

this doesn't give you exactly what you want but you can get some orion
specific stuff from this.

check out the other stuff in the administration.* package.

greg

- Original Message -
From: "Roozbeh Ghaffari" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Thursday, July 26, 2001 2:57 PM
Subject: Finding Default DataSource in My App


> Dear Friends,
>
> Can anybody tell me how to find the default datasource of my application
at
> run-time? (JNDI name or instance ref.) And a more general question: How
can
> I query for Orion specific application information during run-time? I saw
> nothing in Orion API docs.
>
> Thank you in advance,
> Roozbeh/
>
>
> _
> Do You Yahoo!?
> Get your free @yahoo.com address at http://mail.yahoo.com
>
>
>

Re: JMS TopicSubscriber blocks on receive()

[Greg Matthews]

 
johan, thanks for the response.
 
As you can see from my code snippet, a message is 
created and posted, and there is a 5 second pause between publishing the 
message, and then attempting to receive it.
 

In my experiments with a Queue, the message was 
available to a QueueReceiver almost straight away. Is there some difference 
between Queue and Topic based messaging where Topic messages take much longer to 
become available?
 
regards,
greg.

  - Original Message - 
  From: 
  Johan Fredriksson 

  To: Orion-Interest 
  Sent: Wednesday, July 25, 2001 5:20 
  PM
  Subject: Re: JMS TopicSubscriber blocks 
  on receive()
  
  It is quite obvious it is waiting for a message. 
  
   
  So my advice is that you write a client / jsp 
  page / whatever with a TopicBrowser(?) to make sure your message really is 
  there.
   
  Johan 
  
- Original Message - 
From: 
Greg 
Matthews 
To: Orion-Interest 
Sent: Wednesday, July 25, 2001 2:49 
AM
Subject: JMS TopicSubscriber blocks on 
receive()

 
dear all,
 
i've having trouble with a subscriber blocking 
on a receive message call.
 
for the time being i can use a Queue instead of 
a Topic but would like
to know if anyone can spot any stupid error i'm 
making.
 
--- code for publisher 
 
Topic t = (Topic)ServerUtil.get( 
"jms/configTopic" );TopicConnectionFactory tcf = 
(TopicConnectionFactory)ServerUtil.get( "jms/TopicConnectionFactory" 
);TopicConnection tc = tcf.createTopicConnection();TopicSession ts = 
tc.createTopicSession(false,TopicSession.AUTO_ACKNOWLEDGE);
TopicPublisher tp = 
ts.createPublisher(t);
Message m = ts.createTextMessage("hello 
world");
tp.publish(m);
 
// create subscriber in a new 
thread.
new Thread( new 
Listener(tc,ts,t)).start();
 
 code for subscriber 
 
class Listener implements Runnable 
{  TopicConnection tc; TopicSession 
ts; Topic t;
 
 Listener(TopicConnection tc, TopicSession 
ts, Topic t) throws Exception {   this.tc= 
tc;  this.ts = ts;  this.t = 
t; } public void run() 
{try 
{   TopicSubscriber sub = 
ts.createSubscriber(t,null,true);   System.out.println("receiving 
message..." 
);   tc.start();   synchronized(this) 
{wait(5000);   }   System.out.println(sub.receive());    
    // < this line blocks. why?
 
  } catch( Exception e ) 
{
 
   e.printStackTrace();  }  System.out.println("end"); }}
 
thanks,
greg.
 

JMS TopicSubscriber blocks on receive()

[Greg Matthews]

 
dear all,
 
i've having trouble with a subscriber blocking on a 
receive message call.
 
for the time being i can use a Queue instead of a 
Topic but would like
to know if anyone can spot any stupid error i'm 
making.
 
--- code for publisher 
 
Topic t = (Topic)ServerUtil.get( "jms/configTopic" 
);TopicConnectionFactory tcf = (TopicConnectionFactory)ServerUtil.get( 
"jms/TopicConnectionFactory" );TopicConnection tc = 
tcf.createTopicConnection();TopicSession ts = 
tc.createTopicSession(false,TopicSession.AUTO_ACKNOWLEDGE);
TopicPublisher tp = 
ts.createPublisher(t);
Message m = ts.createTextMessage("hello 
world");
tp.publish(m);
 
// create subscriber in a new thread.
new Thread( new 
Listener(tc,ts,t)).start();
 
 code for subscriber 
 
class Listener implements Runnable 
{  TopicConnection tc; TopicSession 
ts; Topic t;
 
 Listener(TopicConnection tc, TopicSession ts, 
Topic t) throws Exception {   this.tc= 
tc;  this.ts = ts;  this.t = 
t; } public void run() 
{try 
{   TopicSubscriber sub = 
ts.createSubscriber(t,null,true);   System.out.println("receiving 
message..." 
);   tc.start();   synchronized(this) 
{wait(5000);   }   System.out.println(sub.receive());    
    // < this line blocks. why?
 
  } catch( Exception e ) {
 
   e.printStackTrace();  }  System.out.println("end"); }}
 
thanks,
greg.
 

JMS: Queue or Topic multicast?

[Greg Matthews]

dear all,
 
are either of the Queue or Topic parts of the JMS 
multicast? i.e. once one subscriber pulls the message off the queue, can other 
subscribers see the message?
 
since each of our clustered servers cache some 
info, i was thinking that a half-reasonable way 
of dealing with some stuff being updated on one server would be for the updated 
server to fire an "i've been updated" message, causing the other servers to 
refresh their cache.
 
i know i could probably look at entity beans, etc, 
but would just like an answer to my question if anyone knows.
 
thanks very much,
greg.

Re: Clustering..

[Greg Matthews]

Title: RE: Clustering..



good point, but i wouldn't see this as a drawback 
-- it's just how it works.
 
isn't that sort of like saying, if i retrieve a 
value from the database, and change it on the client, then the database isn't 
automatically updated?
 

  - Original Message - 
  From: 
  Marcel Schutte 
  To: Orion-Interest 
  Sent: Tuesday, July 17, 2001 6:01 
PM
  Subject: RE: Clustering..
  
   
  -Original Message-From: [EMAIL PROTECTED] 
  [mailto:[EMAIL PROTECTED]]On Behalf Of Kesav 
  KumarSent: Tuesday, July 17, 2001 12:32 AMTo: 
  Orion-InterestSubject: RE: Clustering..
   
   This mechanism have one 
  drawback:  If you keep an object in session and later some place you get 
  the reference of the object and modify the object the modifications won't get 
  reflect across islands:
  Example:     
  UserInfo info = new UserInfo()  // Some kind of 
  object...     info.user_name = "oldName"; 
      session.setAttribute("UserInfo", info) // At this point the info is 
  replicated to all cluster islands 
  //Later at any place of your code 
      UserInfo info = 
  session.getAttribute("UserInfo"); 
      info.user_name = 
  "newName"; //This change won't replicated 
  //In some other program 
      UserInfo info = 
  session.getAttribute("UserInfo"); 
      System.out.println(info.user_name)  // If the request comes to the 
  same cluster where the modification occures you will get the "NewName" if the 
  request goes to another cluster you will get "oldname"
  Work Around:  If you do any kind of modifications to your 
  session objects put the objects again into session using 
  setSessionAttribute.  Keep one thing in mind that when ever you call this 
  setSessionAttribute method the session info is going to replicate across 
  different islands.
  I've noticed this too, is this according to 
  the jsp/servlet specifications or just container specific? When you are using 
  JSP's it is very awkward having to call setAttribute() after every 
   or 
  similar.
  This is the same problem which I faced and lead me to write my 
  own session management. 
  Kesav Kumar Kolla Voquette Inc 
  650 356 3740(W) 510 889 
  6840(R) Voquette...Delivering Sound Information 
  
  -Original Message- From: 
  [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On 
  Behalf Of Duffey, Kevin Sent: Monday, July 16, 2001 
  12:39 PM To: Orion-Interest Subject: Clustering.. 
  Hi all, 
  I know how to cluster Orion..after all, its pretty easy. What 
  I want to know from someone who knows..is if in the 
  1.5.2 version all the bugs (if there were actually 
  any) are worked out. Does Orion cluster with no problems? Does 
  session fail-over and application context fail-over work 
  without a hitch (providing all objects in the session 
  and/or context are serializable)? If you have two 
  machines in an island, and the session begins on one..does it automatically "copy" any session info to the other box for you? If you 
  shut that box off, and all requests go to the 2nd box, 
  does the session still exist with all objects still in 
  memory? 
  I take it each and every request on a clustered box requires 
  the session to be duplicated exactly as is to the 
  other machine? If this is the case, at what "rate" 
  does Orion copy out the session and anything else to another server in the island (or to each and every server in the island)? How 
  does this affect performance as compared to using just 
  a single box without clustering..is it 2 to 3 times 
  slower because of not only copying its session to one 
  or more boxes, but also having to deal with "merging" other boxes context/session information into its own? Or does a cluster 
  simply create the objects and session stuff on each 
  box as its generated (thus..its not some background 
  thread that copies session info when it has time to other boxes)? 
  Is it equally good at clustering on the EJB tier (so 
  scalability on the EJB tier or the front-end WEB tier 
  is equally as easy)? 
  Thanks. 

Re: Application Name

[Greg Matthews]

javax.naming.Context ctx = new InitialContext();
   ApplicationAdministrator admin =
(ApplicationAdministrator)ctx.lookup("java:comp/Administrator");

then print out admin.getDisplayName( ) might be what you're after ???


- Original Message -
From: "Wayne Franklin" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, July 14, 2001 5:30 AM
Subject: Application Name


> Is there a way to query the application name from JSP?
>
> I apologize if this message is a duplicate.  I sent one several hours ago
> and it hasn't showed up yet.  I don't know if this is normal for list
> servers.
>
> Wayne Franklin
> Marathon Innovations, Inc.
>
>
>
>

Re: Orion hangs when connections exhausted.

[Greg Matthews]

milton,

can you post a small code snippet of
- obtaining a reference to the datasource
- getting the connection
- using, then releasing the connection?

the only other sort of thing in the easy to spot category
is that one or more of your methods aren't committing
or rolling back, and thereby leaving database locks
lying around.

does oracle have any tools to check which sql statement
is blocking?

greg.

- Original Message -
From: "Milton S" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Tuesday, July 03, 2001 9:54 AM
Subject: Re: Orion hangs when connections exhausted.


> And the obvious answer is yes.
>
> Greg Matthews wrote:
>
> > just to get the obvious things out of the way first, are you calling
> > connection.close() prior to the EJB method returning?
> >
> > doing this in a finally block is a good idea.
> >
> > - Original Message -
> > From: "Milton S" <[EMAIL PROTECTED]>
> > To: "Orion-Interest" <[EMAIL PROTECTED]>
> > Sent: Tuesday, July 03, 2001 5:36 AM
> > Subject: Orion hangs when connections exhausted.
> >
> > > Has anyone experienced Orion hanging when all of the available
> > > connections from the pool are exhausted? I noticed a similar question
> > > posted by Joel Shellman back in August of 2000 but there was no
response
> > > to his question.
> > >
> > > I can re-issue the same URL from a browser multiple times and once the
> > > application has taken the last pooled connection it hangs and never
> > > returns. Orion never lets go of the connection even after waiting for
> > > over an hour. Only way to get Orion working again is to boot.
> > > Connections are not being made in an EJB but in "regular" java classes
> > > that are called from a stateless session bean.
> > >
> > > My application is to be deployed on an Intranet and I expect the
traffic
> > > to be minimal so a work around is to set up the connection pool with a
> > > large enough number of connections so that they are never used up, but
> > > Orion shouldn't be hanging. Is this an Orion problem or am I missing
> > > something in the data-sources.xml file?
> > >
> > > Thanks for any help.
> > >
> > > Milton S.
> > >
> > > I am running Orion 1.5.2 on NT 2000 Professional against an Oracle 8i
> > > database on NT 4.0 server using the Oracle Thin JDBC driver.
> > >
> > > Here is the data-source I am using:
> > >
> > >  > >  class="com.evermind.sql.DriverManagerDataSource"
> > >  name="jdbc/PWOW_OracleCoreDS"
> > >  location="jdbc/PWOW_OracleCoreDS"
> > >  pooled-location="jdbc/PWOW_OraclePooledDS"
> > >  xa-location="jdbc/xa/PWOW_OracleXaDS"
> > >  ejb-location="jdbc/PWOW_OracleEjbDS"
> > >  connection-driver="oracle.jdbc.driver.OracleDriver"
> > >  min-connections="5"
> > >  max-connections="10"
> > >  username="uid"
> > >  password="pwd"
> > >  url="jdbc:oracle:thin:@:1521:"
> > >  inactivity-timeout="10"
> > >  connection-retry-interval="10"
> > >  wait-timeout="300"
> > > />
> > >
> > >
> > >
> > >
>
>
>

Re: Orion hangs when connections exhausted.

[Greg Matthews]

just to get the obvious things out of the way first, are you calling
connection.close() prior to the EJB method returning?

doing this in a finally block is a good idea.

- Original Message -
From: "Milton S" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Tuesday, July 03, 2001 5:36 AM
Subject: Orion hangs when connections exhausted.


> Has anyone experienced Orion hanging when all of the available
> connections from the pool are exhausted? I noticed a similar question
> posted by Joel Shellman back in August of 2000 but there was no response
> to his question.
>
> I can re-issue the same URL from a browser multiple times and once the
> application has taken the last pooled connection it hangs and never
> returns. Orion never lets go of the connection even after waiting for
> over an hour. Only way to get Orion working again is to boot.
> Connections are not being made in an EJB but in "regular" java classes
> that are called from a stateless session bean.
>
> My application is to be deployed on an Intranet and I expect the traffic
> to be minimal so a work around is to set up the connection pool with a
> large enough number of connections so that they are never used up, but
> Orion shouldn't be hanging. Is this an Orion problem or am I missing
> something in the data-sources.xml file?
>
> Thanks for any help.
>
> Milton S.
>
> I am running Orion 1.5.2 on NT 2000 Professional against an Oracle 8i
> database on NT 4.0 server using the Oracle Thin JDBC driver.
>
> Here is the data-source I am using:
>
>   class="com.evermind.sql.DriverManagerDataSource"
>  name="jdbc/PWOW_OracleCoreDS"
>  location="jdbc/PWOW_OracleCoreDS"
>  pooled-location="jdbc/PWOW_OraclePooledDS"
>  xa-location="jdbc/xa/PWOW_OracleXaDS"
>  ejb-location="jdbc/PWOW_OracleEjbDS"
>  connection-driver="oracle.jdbc.driver.OracleDriver"
>  min-connections="5"
>  max-connections="10"
>  username="uid"
>  password="pwd"
>  url="jdbc:oracle:thin:@:1521:"
>  inactivity-timeout="10"
>  connection-retry-interval="10"
>  wait-timeout="300"
> />
>
>
>
>

Re: Question about port listener

[Greg Matthews]

yes, but theoretically, EJBs aren't supposed to interact with the real
world, i.e. they're not supposed to manipulate files, open sockets, or do
anything useful unless it's through the container's services.

you're choices are:
1. do what you said, i.e. open a socket. it would be against the spec but
who cares if it works right?
2. if everyone's on the same network, look at using JMS, and then getting
your client to put a message in the JMS
3. open the socket from a servlet, not a bean, which is sort of the same as
option 1 but within the EJB spec.


- Original Message -
From: "Karri Niemelä" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, June 30, 2001 3:03 AM
Subject: Question about port listener


> Question 1) Is it possible to build an bean, which is some sort of tcp
port
> listener?
> Idea is to listen for traffic on port and then fire an even/send a msg to
> queue if something is received
>
> Question 2) What if we have an 3rd party java api to connect some server
> using rmi
> (for an example an api which builds up an connection to short message
> center)
> Would it be possible to create an bean which builds up this connection(and
> does this only once),
> keeps it up and listens for method calls from clients and then sends them
> away
>
>
>
>

Re: clustering + ssl together

[Greg Matthews]

cool. thanks for the info. we're using windows
so i'll try to track down a version.

look like you've nailed the bug? down pretty well.

hopefully, the orion lads will be nice enough to fix
it for the next build.

- Original Message -
From: "elephantwalker" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Friday, June 29, 2001 8:59 AM
Subject: RE: clustering + ssl together


> Greg,
>
> I found a tool that you can use to look at what's going on with ssl and
the
> loadbalancer or orion. If you are using linux, you probably have this
> already, openssl. I am not sure if there is a windows build, though.
>
> at the command line:
>
> openssl s_client -connect loadbalancermachine:443 -state -debug
>
> The result is clear, ssl returns with an ssl handshake failure. If you do
> this on the an instance of orion with ssl:
>
> openssl s_client -connect orionmachine:443 -state -debug
>
> There is no problem.
>
> So that nails the bug down, its got to be the loadbalancer. I will post
this
> in bug 525 for Karl.
>
> Regards,
>
> the elephantwalker
>
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of elephantwalker
> Sent: Thursday, June 28, 2001 2:22 PM
> To: Orion-Interest
> Subject: RE: clustering + ssl together
>
>
> Greg,
>
> I just tried this...
>
> 1. I assummed that ssl was completely broken for the loadbalancer. So I
> stripped off the secure=true from my loadbalancer on port 443 and all of
my
> secure-web-site.xml's backends'.
>
> 2. I created a new orion instance with only the secure-web-site.xml in the
> server.xml.
>
> 3. I modified the global-web-application.xml so that the only servlet and
> mapping is this:
>
>   
>tunnel
>com.evermind.server.http.TunnelServlet
>
> targetRoot
> http://loadbalancermachine:443/
>
>   
>   
>tunnel
>/*
>   
>
> 5. I started the new "proxy" orion instance.
>
> With my browser, I put https://proxymachine/
>
> ...
>
> Voila! it worked!
>
> Its a little slow, so you could probably do this with a reverse-proxy, ssl
> apache to get faster response.
>
> This is only a workaround, since the loadbalancer ssl is broken now.
>
> Regards,
>
> the elephantwalker
>
> .ps the only caveat here is the j2ee j_security_check won't work, but
> otherwise, everything works. I don't understand why, though.
>
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> Sent: Wednesday, June 27, 2001 12:00 AM
> To: Orion-Interest
> Subject: Re: clustering + ssl together
>
>
> ew,
>
> using your example, i have tried the equivalent of
> https://localhost/mysecuresite/login
> which should have gone to port 443.
>
> in the effort to "look under orion's covers".
>
> i've seen a -Djavax.net.debug=all flag mentioned in a previous post by
> tomas anderson (27.6.01), and gave it a try but no extra output
> appeared in orion. do you know what this is supposed to show?
>
> do you know if there is a way to see where the request is getting up to?
> can we do a netstat or something to see where the request is falling over
> or what processes are listening on what ports?
>
> greg.
>
> - Original Message -
> From: "elephantwalker" <[EMAIL PROTECTED]>
> To: "Orion-Interest" <[EMAIL PROTECTED]>
> Sent: Wednesday, June 27, 2001 2:58 PM
> Subject: RE: clustering + ssl together
>
>
> > Greg,
> >
> > I just tried something which ALMOST worked. I tried the secure
> loadbalancer
> > instance like this in the browser:
> >
> > http://localhost:443/mysecuresite/login.
> >
> > The secure loadbalancer showed a session id, and forwarded the request
to
> > the secure island! Of course the site didn't do anything, since it was
> > looking for a handshake. It looks like the loadbalancer is just not
doing
> > its bit...it is refusing all connections which are secure.
> >
> > regards,
> >
> > the elephantwalker
> >
> >
> > -Original Message-
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> > Sent: Tuesday, June 26, 2001 3:00 PM
> > To: Orion-Interest
> > Subject: Re: clustering + ssl together
> >
> >
> >
> > ew,
> >
> > i was trying to run a single secure load balancer
> > with it's own load-balancer.xml.
> >
> > loadbalancer did register the 2

Re: clustering + ssl together

[Greg Matthews]

nice one.

since the purpose of the work we've both been doing
is to get a clustered SSL setup going, would it be
worth adding something to the bug report?

i.e. which asks karl and magnus very nicely to see if they
can create a clustered SSL setup?

greg

- Original Message -
From: "elephantwalker" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, June 27, 2001 3:57 PM
Subject: RE: clustering + ssl together


> Greg,
>
> I just logged this as bug 525. The ssl loadbalancer just won't accept
> connections with https://, but will accept connections with http://. Basic
> problem with the code. Its not us. Karl and Magnus need to fix this.
>
> regards,
>
> the elephantwalker
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of elephantwalker
> Sent: Tuesday, June 26, 2001 9:59 PM
> To: Orion-Interest
> Subject: RE: clustering + ssl together
>
>
> Greg,
>
> I just tried something which ALMOST worked. I tried the secure
loadbalancer
> instance like this in the browser:
>
> http://localhost:443/mysecuresite/login.
>
> The secure loadbalancer showed a session id, and forwarded the request to
> the secure island! Of course the site didn't do anything, since it was
> looking for a handshake. It looks like the loadbalancer is just not doing
> its bit...it is refusing all connections which are secure.
>
> regards,
>
> the elephantwalker
>
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> Sent: Tuesday, June 26, 2001 3:00 PM
> To: Orion-Interest
> Subject: Re: clustering + ssl together
>
>
>
> ew,
>
> i was trying to run a single secure load balancer
> with it's own load-balancer.xml.
>
> loadbalancer did register the 2 orions i'd set up to appear
> in the cluster, but after being able to see them appear on
> the loadbalancer screen, i was still unable to access my
> web app. the browser just sat there with the little IE
> symbol spinning, but no joy.
>
> all orions and the loadbalancer had their own keystore
> setup using a test certificate generated from thawte.com
>
> loadbalancer => secure and on port 443 (on box1)
> orion1 => secure and on port 443 (on box2)
> orion2 => secure and on port 8080 (on box1) !! but only in some
experiments.
>
> i also tried various other configurations of the loadbalancer
> and cluster machines having secure on/off, etc. and
> swapping the port numbers around, e.g. when loadbalancer
> and orion2 were both running, they were both secure="true"
> but obviously only one can run on port 443 at one time, so
> i made orion2 run on port 8080 while secure="true" was set.
>
> i also had a look at apache for how to setup SSL but it looks
> like you've got to compile the mod in yourself for win32 so
> i've given that a miss for the moment.
>
> greg.
>
> - Original Message -
> From: "elephantwalker" <[EMAIL PROTECTED]>
> To: "Orion-Interest" <[EMAIL PROTECTED]>
> Sent: Wednesday, June 27, 2001 2:48 AM
> Subject: RE: clustering + ssl together
>
>
> > Here are the  in the plan so far...see below.
> >
> >
> > -Original Message-
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]]On Behalf Of elephantwalker
> > Sent: Monday, June 25, 2001 1:29 AM
> > To: Orion-Interest
> > Subject: RE: clustering + ssl together
> >
> >
> > Greg,
> >
> > I am doing this now, so I will get back to the list when I am finished.
> This
> > is my working plan:
> >
> > 1. there are two loadbalancers instances, one for http and one for
https.
> > These can be on the same machine or seperate machines.
> >
> > 
> >
> > At one level this works, but you have to set the
> minimumIsland/maximumIsland
> > so that each respective loadbalancer picks up either the https island or
> the
> > http island. However, https connections do not work. It could be because
> of
> > this blurb in the load-balancer.xml description:
> >
> > secure - Whether or not to use SSL. The default is false. SSL is only
used
> > when using session (not IP)
> >based balancing and the backend and the site is using SSL. If you
> specify
> > the balancer to use SSL then
> >the backend servers will not (the balancer converts to HTTP, ie
> contains
> > the SSL layer). Note that this
> >puts the strain of decoding the SSL on the balancer.
> >
> > I'm sorry, but does this say that we have the op

Re: clustering + ssl together

[Greg Matthews]

ew,

using your example, i have tried the equivalent of
https://localhost/mysecuresite/login
which should have gone to port 443.

in the effort to "look under orion's covers".

i've seen a -Djavax.net.debug=all flag mentioned in a previous post by
tomas anderson (27.6.01), and gave it a try but no extra output
appeared in orion. do you know what this is supposed to show?

do you know if there is a way to see where the request is getting up to?
can we do a netstat or something to see where the request is falling over
or what processes are listening on what ports?

greg.

- Original Message -
From: "elephantwalker" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, June 27, 2001 2:58 PM
Subject: RE: clustering + ssl together


> Greg,
>
> I just tried something which ALMOST worked. I tried the secure
loadbalancer
> instance like this in the browser:
>
> http://localhost:443/mysecuresite/login.
>
> The secure loadbalancer showed a session id, and forwarded the request to
> the secure island! Of course the site didn't do anything, since it was
> looking for a handshake. It looks like the loadbalancer is just not doing
> its bit...it is refusing all connections which are secure.
>
> regards,
>
> the elephantwalker
>
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> Sent: Tuesday, June 26, 2001 3:00 PM
> To: Orion-Interest
> Subject: Re: clustering + ssl together
>
>
>
> ew,
>
> i was trying to run a single secure load balancer
> with it's own load-balancer.xml.
>
> loadbalancer did register the 2 orions i'd set up to appear
> in the cluster, but after being able to see them appear on
> the loadbalancer screen, i was still unable to access my
> web app. the browser just sat there with the little IE
> symbol spinning, but no joy.
>
> all orions and the loadbalancer had their own keystore
> setup using a test certificate generated from thawte.com
>
> loadbalancer => secure and on port 443 (on box1)
> orion1 => secure and on port 443 (on box2)
> orion2 => secure and on port 8080 (on box1) !! but only in some
experiments.
>
> i also tried various other configurations of the loadbalancer
> and cluster machines having secure on/off, etc. and
> swapping the port numbers around, e.g. when loadbalancer
> and orion2 were both running, they were both secure="true"
> but obviously only one can run on port 443 at one time, so
> i made orion2 run on port 8080 while secure="true" was set.
>
> i also had a look at apache for how to setup SSL but it looks
> like you've got to compile the mod in yourself for win32 so
> i've given that a miss for the moment.
>
> greg.
>
> - Original Message -
> From: "elephantwalker" <[EMAIL PROTECTED]>
> To: "Orion-Interest" <[EMAIL PROTECTED]>
> Sent: Wednesday, June 27, 2001 2:48 AM
> Subject: RE: clustering + ssl together
>
>
> > Here are the  in the plan so far...see below.
> >
> >
> > -Original Message-
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]]On Behalf Of elephantwalker
> > Sent: Monday, June 25, 2001 1:29 AM
> > To: Orion-Interest
> > Subject: RE: clustering + ssl together
> >
> >
> > Greg,
> >
> > I am doing this now, so I will get back to the list when I am finished.
> This
> > is my working plan:
> >
> > 1. there are two loadbalancers instances, one for http and one for
https.
> > These can be on the same machine or seperate machines.
> >
> > 
> >
> > At one level this works, but you have to set the
> minimumIsland/maximumIsland
> > so that each respective loadbalancer picks up either the https island or
> the
> > http island. However, https connections do not work. It could be because
> of
> > this blurb in the load-balancer.xml description:
> >
> > secure - Whether or not to use SSL. The default is false. SSL is only
used
> > when using session (not IP)
> >based balancing and the backend and the site is using SSL. If you
> specify
> > the balancer to use SSL then
> >the backend servers will not (the balancer converts to HTTP, ie
> contains
> > the SSL layer). Note that this
> >puts the strain of decoding the SSL on the balancer.
> >
> > I'm sorry, but does this say that we have the option of NOT using SSL
for
> > the balancer, but using it for the backend? Or if we use SSL for the
> > balancer, SSL isn't used on the backend (and thus we have to strip all
of
> > t

Re: clustering + ssl together

[Greg Matthews]

hey
> have so much invested in the "apache/oracle" solution.
>
> 
>
>  This option is looking better and better.
>
> 
>
> I'm testing this now, as soon as I get through the hickups, I will let the
> list know.
>
> regards,
>
> the elephantwalker
>
>
>
>
>
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> Sent: Sunday, June 24, 2001 3:02 PM
> To: Orion-Interest
> Subject: clustering + ssl together
>
>
>
> dear all,
>
> there has been a recent post on this but no solution posted.
> i've got some more info on the problem.
>
> can the developers of orion or anyone else let me know
> if anyone has successfully set up an ssl orion cluster?
>
> i can:
> - set up clustering
> - set up ssl
>
> ...but not both together.
>
> some clues.
>
> 1. on orionserver.com there is doco for load-balancer.xml that
> suggests loadbalancer.jar can be given SSL keystore information.
> does this mean that a clustered SSL setup requires loadbalancer
> to share the same keystore as each box in the cluster?
>
> 2. how do you set the web-site.xml for a clustered secure app.
>
> you can't have both the loadbalancer + your secure app
> both running on port 443 on the same box, so what do you
> do?
> i) run loadbalancer on another port?
> ii) run your app on another port?
> - the orion doco says that when your app needs to
>   be made secure you should add a secure="true"
>   attribute to the web-site element of the web-site.xml
>   plus remove the port attribute.
>
> if someone has made this work i'd be grateful for any information,
> or if you couldn't be bothered explaining how to do it, just maybe
> forward me your server.xml, loadbalancer.xml, web-site.xml and
> i'll work it out from that.
>
> thanks.
> greg.
>
>
>

clustering + ssl together

[Greg Matthews]

 
dear all,
 
there has been a recent post on this but no 
solution posted.
i've got some more info on the 
problem.
 
can the developers of orion or anyone else let me 
know
if anyone has successfully set up an ssl orion 
cluster?
 
i can:
- set up clustering
- set up ssl
 
...but not both together.
 
some clues.
 
1. on orionserver.com there is doco for 
load-balancer.xml that
    suggests loadbalancer.jar can be 
given SSL keystore information.
    does this mean that a clustered 
SSL setup requires loadbalancer
    to share the same keystore as 
each box in the cluster?
 
2. how do you set the web-site.xml for a clustered 
secure app.
 
    you can't have both the 
loadbalancer + your secure app
    both running on port 443 on the 
same box, so what do you
    do?
        i) run 
loadbalancer on another port?
        ii) run your 
app on another port?
        
    - the orion doco says that when your app needs to 

  be 
made secure you should add a secure="true"
  attribute 
to the web-site element of the web-site.xml
  plus 
remove the port attribute.
 
if someone has made this work i'd be grateful for 
any information,
or if you couldn't be bothered explaining how to do 
it, just maybe
forward me your server.xml, loadbalancer.xml, 
web-site.xml and
i'll work it out from that.
 
thanks.
greg.

Re: RMI Clustering done. Now want more.

[Greg Matthews]

hi.

not really.

i just wondered if you knew what the clustering settings in rmi.xml are for.

i said that i'd sort of got clustering working, and hadn't added any rmi.xml
clustering settings, so what then do they do?

thanks,
greg.

- Original Message -
From: "Lachezar Dobrev" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Monday, June 18, 2001 4:57 PM
Subject: Re: RMI Clustering done. Now want more.


> > hi lachezar,
>Hya.
>Looking at the text I think you misunderstood me.
>I was not looking for a fail-over or Web clustering solution. That's
it.
>I need to use Orion as a EJB server, not the fulll J2EE App server.
>I have some QUITE heavy beans, and just very small web-interface for
more
> or less a couple of users. I don't have a
> super-duper-overloaded-with-connections site. Instead I have
time-consuming
> stateless session beans, that take up to a 100% CPU time, and work for
some
> couple of hours.
>So what I need is NOT a WEB-Failover solution, but rather a RMI
> (EJB-to-EJB) clustering in order to put my heavy-handed EJBs allover my
> intranet.
>
>Hope I clarified this.
>
>BTW. I forgot to mention something:
>LINUX ISSUE! (tested on RH7/JDK1.3.1)
> Define a default route in the routing table.
> Otherwise setting the multicast cluster fails with an exception.
> This might be as well mentioned on OrionSupport or OrionServer.
> THIS IS A JVM ISSUE. NOT ORION SPECIFIC!
>
>Lachezar Dobrev
>Software developer
>LSN Software and Telecommunication services
>Sofia, Bulgaria
>
> > i've got clustering working for our app but haven't used the rmi.xml
> > clustering settings. the app seems to failover ok to another machine in
> the
> > cluster if i switch a machine off, etc,.
> >
> > we use only stateless session beans and so i wondered if only entity
bean
> > information was exchanged over the port specified in rmi.xml
> >
> > what's all this stuff for in rmi.xml
> >
> >  > username="cluster-user" />
> > Tag that is defined if the application is to be clustered. Used to set
up
> a
> > local multicast cluster. A username and password used for the servers to
> > intercommunicate also needs to be specified.
> >
> > host - The multicast host/ip to transmit and receive cluster data on.
The
> > default is 230.0.0.1.
> > id - The id (number) of this cluster node to identify itself with in the
> > cluster. The default is based on local machine IP.
> > password - The password configured for cluster access. Needs to match
that
> > of the other nodes in the cluster.
> > port - The port to transmit and receive cluster data on. The default is
> > 9127.
> > username - The username configured for cluster access. Needs to match
that
> > of the other nodes in the cluster.
> >
> > thanks,
> > greg.
> >
> >
> > - Original Message -
> > From: Lachezar Dobrev
> > To: Orion-Interest
> > Sent: Friday, June 15, 2001 6:18 PM
> > Subject: RMI Clustering done. Now want more.
> >
> >
> >Hello.
> >OK. I was able to (at last) run the RMI clustering. It now works.
> But...
> > I have a question...
> >Now I can put up a cluster, and when a bean does not exist on the
> server
> > it is searched on the cluster and used. However if it is deployed on
more
> > than one machine... It is not found on all of them, but rather on only
> one.
> >
> >For simplicity let's assume we have OrionA, OrionB, OrionC and
OrionD.
> >a) on all machines we have EnterpriseBean1 deployed. It accesses
> > EnterpriseBean2
> >b) machine OrionA does not have EnterpriseBean2 deployed.
> >c) machines OrionB, OrionC and OrionD have EnterpriseBean2 deployed.
> >d) all machines have a RMI cluster set-up and working.
> >   (BTW) Because I needed quite some time to understand why the
> >   cluster was not working I am eager to help anyone, that needs
> >   help on the subject.
> >
> >If machine OrionA runs EnterpriseBean1 it will have to look-up
> > EnterpriseBean2, which is used by EnterpriseBean1.
> >There is no problem to find that bean (it exists on the cluster).
> > However... Once EnterpriseBean2 is used on OrionC it keeps using that
one.
> > Why? Also. If we deploy EnterpriseBean2 on OrionA it is always found
there
> > and not on ANY other cluster machine.
> >
> >1. How does Orion know on which machine to use the bean? The last
one,
> > that wakes up?
> >2. Is there a way to implement (at least) a round-robin technique to
> > access the clustered bean on more than one server? I mean... When I
> look-up
> > EnterpriseBean2... I would like to have a different server answering
every
> > time. Round-robin is sufficient, even though a loadbalancing technique
> would
> > be better appropriate.
> >
> >Ok. Waiting forward to hearing from you all...
> >
> >P.S. Another question... CAN I TURN WEB-SERVICES OFF??? I tried
> removing
> > the default-web-app tag in the server.xml, but it starts complaining
about
> 

.ear files + library tags + DB drivers

[Greg Matthews]

i've sort of had success with this but been unable 
to install the
database driver jar file using this 
technique.
 
i suspect it's a classloader order 
problem.
 
i receive an error message from orion
 
"Error initializing data-source . 
'com.inet.tds.TdsDriver' not found.
 
1. when i put Una2000.jar (the db driver jar) in 
the orion lib directory
    i receive no error 
messages.
 
2. the other jars that i have bundled with the .ear 
file seem to be
    accessible through the 
 entry i created in the
    orion-application.xml 
file.
 
is there some trick in getting the library stuff 
processed before
orion starts checking datasources, 
etc.
 
thanks,
greg.
 

  - Original Message - 
  From: 
  Tim Drury 
  
  To: Orion-Interest 
  Sent: Friday, June 08, 2001 11:20 
PM
  Subject: RE: simplifying EAR 
  deployment
  
   
  Greg,
   
  put 
  the 3rd party jars in $EAR_ROOT/lib then edit your 
  orion-application.xml
  file 
  (in $EAR_ROOT/META-INF or $EAR_ROOT/orion for slightly older 
  versions
  of 
  orion) adding: .  If you don't have 
  orion-application.xml,
  deploy your app once and grab the auto-generated copy from 
  
  orion/application-deployment/yourapp.  put the copy in your source 
  tree, edit it, 
  and 
  add it to your ear for subsequent deployments.
   
  -tim
  efnet #java joeblowgt
  
-----Original Message-From: Greg Matthews 
[mailto:[EMAIL PROTECTED]]Sent: Thursday, June 07, 2001 8:44 
PMTo: Orion-InterestSubject: simplifying EAR 
deployment
dear all,
 
is there any way to have 3rd party jar files 
included in an EAR then deployed into the orion\lib directory.
 
i'm trying to minimise the number of steps our 
clients have to perform when installing our app.
 
thanks,
greg

Re: RMI Clustering done. Now want more.

[Greg Matthews]

hi lachezar,

i've got clustering working for our app but haven't used the rmi.xml
clustering settings. the app seems to failover ok to another machine in the
cluster if i switch a machine off, etc,.

we use only stateless session beans and so i wondered if only entity bean
information was exchanged over the port specified in rmi.xml

what's all this stuff for in rmi.xml


Tag that is defined if the application is to be clustered. Used to set up a
local multicast cluster. A username and password used for the servers to
intercommunicate also needs to be specified.

host - The multicast host/ip to transmit and receive cluster data on. The
default is 230.0.0.1.
id - The id (number) of this cluster node to identify itself with in the
cluster. The default is based on local machine IP.
password - The password configured for cluster access. Needs to match that
of the other nodes in the cluster.
port - The port to transmit and receive cluster data on. The default is
9127.
username - The username configured for cluster access. Needs to match that
of the other nodes in the cluster.

thanks,
greg.


- Original Message -
From: Lachezar Dobrev
To: Orion-Interest
Sent: Friday, June 15, 2001 6:18 PM
Subject: RMI Clustering done. Now want more.


   Hello.
   OK. I was able to (at last) run the RMI clustering. It now works. But...
I have a question...
   Now I can put up a cluster, and when a bean does not exist on the server
it is searched on the cluster and used. However if it is deployed on more
than one machine... It is not found on all of them, but rather on only one.

   For simplicity let's assume we have OrionA, OrionB, OrionC and OrionD.
   a) on all machines we have EnterpriseBean1 deployed. It accesses
EnterpriseBean2
   b) machine OrionA does not have EnterpriseBean2 deployed.
   c) machines OrionB, OrionC and OrionD have EnterpriseBean2 deployed.
   d) all machines have a RMI cluster set-up and working.
  (BTW) Because I needed quite some time to understand why the
  cluster was not working I am eager to help anyone, that needs
  help on the subject.

   If machine OrionA runs EnterpriseBean1 it will have to look-up
EnterpriseBean2, which is used by EnterpriseBean1.
   There is no problem to find that bean (it exists on the cluster).
However... Once EnterpriseBean2 is used on OrionC it keeps using that one.
Why? Also. If we deploy EnterpriseBean2 on OrionA it is always found there
and not on ANY other cluster machine.

   1. How does Orion know on which machine to use the bean? The last one,
that wakes up?
   2. Is there a way to implement (at least) a round-robin technique to
access the clustered bean on more than one server? I mean... When I look-up
EnterpriseBean2... I would like to have a different server answering every
time. Round-robin is sufficient, even though a loadbalancing technique would
be better appropriate.

   Ok. Waiting forward to hearing from you all...

   P.S. Another question... CAN I TURN WEB-SERVICES OFF??? I tried removing
the default-web-app tag in the server.xml, but it starts complaining about
no default web-app :((( I want to have some Orion servers to act as EJB only
servers.

   Lachezar

Re: clustering and key generation

[Greg Matthews]

i think elephantwalker's solution is probably better from a
database independence point of view and understand
now that counter works off the db.

i think using identity's is probably ok also, but we'll have
to do some OO stuff server side to abstract away the
"get the last identity used" (MS SQL SERVER) from
"get the next value from the sequence" (ORACLE).

cheers,
greg

- Original Message -
From: "elephantwalker" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Monday, June 11, 2001 1:24 PM
Subject: RE: clustering and key generation


> We have several orion's running in clustering, each of them serving up
keys
> from counter.jar, with no conflicts. Let me be clear, each orion instance
> uses its own local counter.jar ejb to generate a key, but the underlying
> data for the key generation comes from the database. The counter.jar is
> called directly from a slsb, which then passes the generated key to the
> entity bean create method. (We could have done this from within the entity
> bean, it just seemed a little faster from the slsb). The slsb method to
> create the entity bean is called from the web tier, a type II controller
> servlet.
>
> If one orion fails, then the other orion can take over, since the web tier
> is clustered as is the loadbalancer's job. We have tested this
> functionality, and it if we turn off one appserver, the other one takes
over
> the controller session without a problem.
>
> You can directly use the database to create the keys from within the
> controller servlet or a bean instanced in the controller, but then you do
> not have any transaction control or any of those nice pooling things going
> on under the hood with the app server, and you will pay for this with a
> performance slowdown.
>
> Brett McGlauphlin's series of articles on Flashline.com clearly indicates
> the reasons for abstracting the key generation to the enterprise tier. I
> would suggest a good read of this.
>
> Regards,
>
> the elephantwalker
>
>
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Ate Douma
> Sent: Sunday, June 10, 2001 6:05 PM
> To: Orion-Interest
> Subject: Re: clustering and key generation
>
>
> If  I understand Greg's decision correctly, he made it to prevent a single
> point of failure on the Orion server instance serving the key generation
> with the counter.jar. That Orion server indeed is a single point of
failure
> instance as only one server should serve the key generation locally.
> Of course, using the database as single point of failure is just a slight
> shift of focus, but those beasts are usually a bit more stable thus I
think
> his decision probably improved the reliability of his system.
>
> Ate
>
> - Original Message -
> From: "elephantwalker" <[EMAIL PROTECTED]>
> To: "Orion-Interest" <[EMAIL PROTECTED]>
> Sent: Monday, June 11, 2001 00:52
> Subject: RE: clustering and key generation
>
>
> > Greg,
> >
> > I didn't really understand your problem. If you are using counter.jar to
> > generate your keys, then the key is actually generated based upon the
last
> > key in the database, not the appserver, so clustering shouldn't be a
> > problem. If there is an issue with transaction concurrency, you can
always
> > hack the ejb-jar.xml for counter.jar to keep control of the
transactions.
> >
> > If you are using jdbc and a slsb (see Brett McGlauphlin's column on
> > Flashline.com), again the database is keeping track of the keys
generated,
> > and not the appserver, so clustering shouldn't be a problem.
> >
> > If you are using a bean (not an ejb, just anyolbean) to generate your
> keys,
> > then you've got a problem with clustering...forget about this approach,
it
> > will never work.
> >
> > We use counter.jar, and have had no problems with key generation in a
> > clustered environment. Counter.jar is in the newsapp, and is freely
> > available for anybody to use.
> >
> > Regards,
> >
> > the elephantwalker
> >
> >
> > -Original Message-
> > From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> > Sent: Sunday, June 10, 2001 3:19 PM
> > To: Orion-Interest
> > Subject: Re: clustering and key generation
> >
> >
> >
> > jason,
> >
> > thankyou for yor responses.
> >
> > in the interests of "keeping it simple", i've decided to try to lobby
> > the rest of the team to go back to using db generated keys (

Re: clustering and key generation

[Greg Matthews]

jason,

thankyou for yor responses.

in the interests of "keeping it simple", i've decided to try to lobby
the rest of the team to go back to using db generated keys (i.e. identity
columns in the case of ms sql server ) and throw out key our
key generation code.

we'll then have a single .ear that can be deployed on any box
with minimal changes to 1 or 2 orion files to allow clustering, and
the only single point of failure will then be the database, and not
the box generating keys.

cheers,
greg.

- Original Message -
From: "Jason Smith" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Sunday, June 10, 2001 2:36 AM
Subject: RE: clustering and key generation


> Have you tried setting:
> 
> in your orion-application.xml on machines B,C, and D?  The only place the
> KeyGenerator bean is really deployed is on A, so machine A's
> orion-application.xml will have remote="false".  I am assuming you have
> already set up your rmi.xml, etc. correctly to support this kind of
> operation (as in the links I posted earlier).
>
> The only other thing I can think of right now is maybe try making a parent
> application which has the KeyGenerator bean and run children apps on the
> other machines.  I haven't tried the parent/child app deployment, so you
> would have to check the archives to see if this is feasible.
>
> -jason
>
>
>

Re: clustering and key generation

[Greg Matthews]

hi,

i had a read and i'm still stuck.

my question specifically is about why the InitialContext i create
for referencing a KeyGeneratorBean on another box doesn't work.

machines A, B, C, and D each have the same app deployed,
which has many beans. Each box should use it's own beans,
except KeyGeneratorBean.

I want to tell *all* boxes to use the KeyGeneratorBean on machine A.

I do this by creating a new InitialContext that is used only when getting
a home interface to KeyGeneratorBean, by calling the InitialContext(
Hashtable)
constructor, where Hashtable contains

java.naming.factory.initial=com.evermind.server.rmi.RMIInitialContextFactory
java.naming.security.principal=
java.naming.security.credentials=
java.naming.provider.url=ormi://machineA/myApp

this works when i run a standalone dos program to get a reference to
KeyGeneratorBean on machineA, so why doesn't it work when
the same code is run from inside an EJB on machineB?

any help much appreciated,
greg.
- Original Message -
From: "Jason Smith" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Friday, June 08, 2001 2:04 PM
Subject: RE: clustering and key generation


> These posts in the archive may help you (although they target Orion web
> server->Orion ejb server configuration instead of a cluster).
>
> http://www.mail-archive.com/orion-interest@orionserver.com/msg12704.html
> http://www.mail-archive.com/orion-interest@orionserver.com/msg11905.html
>
> -jason
>
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Greg Matthews
> Sent: Thursday, June 07, 2001 10:19 PM
> To: Orion-Interest
> Subject: clustering and key generation
>
>
> dear all,
>
> if there any way to get all machines in a cluster to lookup a stateless
> session bean (KeyGeneratorBean) on *one* of the machines in the cluster
> only.
>
> i've given it a try but can't seem to find a way to get machine B to use
> machine A's KeyGeneratorBean, even though machine B builds a new
> InitialContext with the 4 environment parameters, e.g.
> principal/credentials/url/factory when doing a lookup for KeyBean.
>
> thanks,
> greg.
>
>
>

clustering and key generation

[Greg Matthews]

dear all,
 
if there any way to get all machines in a cluster 
to lookup a stateless session bean (KeyGeneratorBean) on *one* of the machines 
in the cluster only.
 
i've given it a try but can't seem to find a way to 
get machine B to use machine A's KeyGeneratorBean, even though machine B builds 
a new InitialContext with the 4 environment parameters, e.g. 
principal/credentials/url/factory when doing a lookup for KeyBean.
 
thanks,
greg.

simplifying EAR deployment

[Greg Matthews]

dear all,
 
is there any way to have 3rd party jar files 
included in an EAR then deployed into the orion\lib directory.
 
i'm trying to minimise the number of steps our 
clients have to perform when installing our app.
 
thanks,
greg

Re: BIG Problem with SSL

[Greg Matthews]

robert,

check out the shared attribute of a web app in the orion docs.

it's designed allow two different web apps to share sessions.

i.e. in the -web-site.xml file, set


  


for both the secure and non-secure apps.

cheers,
greg

- Original Message -
From: "Robert S. Sfeir" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, May 09, 2001 5:19 AM
Subject: BIG Problem with SSL


> When I setup SSL under Orion and I hit:
> http://www.domain.com
>
> then hit
>
> https://www.domain.com
>
> they both come up fine, but it looks like Orion is starting a new
session!!!
>
> This is bad because if someone is logged in under https, it doesn't look
> like they're logged in when they're under the http server.
>
> Am I missing a setting with SSL or is this normal Orion behavior?
>
> Thanks for any prompt response you might have.
>
> R
>
>
>

java:comp/Administrator

[Greg Matthews]

Dear all,
 
By following some code in a previous post i can get 
a reference to an Administrator object
which lets me do things like get the 
application.xml .
 
ApplicationAdministrator admin = 
(ApplicationAdministrator)ctx.lookup("java:comp/Administrator");

System.out.println( 
admin.getDisplayName());
 
 
Does anyone know what other Orion objects are 
available to bind to?
 
 
As an aside, if i 
try to iterate through the object under java:comp 
with a NamingIterator i 
only get the following, which is strange because i was expecting to see Administrator 
there (?)
 
UserTransaction: 
com.evermind.server.ApplicationServerTransactionManagerRoleManager: 
com.evermind.server.hmLogger: 
com.evermind.util.CollectionLoggerUserManager: 
com.evermind.server.XMLUserManagerenv: com.evermind.naming.j_ThreadPool: 
com.evermind.server.he
 
 
Thanks,
Greg.

Re: ASP config + clustering

[Greg Matthews]

that's fine except if you need to run 2 or more boxes to
handle the load.

i want a single copy of the application specific config files
shared by each clustered orion box.

i also want to be able to programmatically determine where
the config files are.

if each orion box in the cluster knows that the root config
directory is \\configmachine\configs\ then i can append
say, the application name, to this to get a config directory.

e.g.  \\configmachine\configs\client1\config.xml for client 1
   \\configmachine\configs\client2\config.xml for client 2

i can then change the application specific config details in a
single place for client 1, and all of client 1's clustered application
copies running across various boxes get the new information.

the question is, where do i get "client1" or "client2" when building
the path?

the best i've been able to do is follow a previous post and get
the application.xml defined  value, which doesn't
really solve the problem because that's included in the .ear file.

i was hoping to be able to programmatically determine the
server.xml defined application name and use that to build
the path.

greg.

- Original Message -
From: "Frank Eggink" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Thursday, March 22, 2001 7:03 PM
Subject: RE: ASP config + clustering


> Greg,
>
> Why run an instance of orion per client you are serving? I'm looking to
> host an applications for a number of clients too, but I'm looking into
> another route.
>
> We plan to 'copy' the applications and run them from one server-instance.
> Guess for resource usage we are better off, just copying the applications
> and have your clients served by one orion instance. Clients would connect
> like //www.host.com/clientname/ or something like that.
>
> Additionally you can (on UNIX/LINUX) soft-link the original
application.ear
> instead of copying it.
>
> As long as you write the apps for the server, things should be secure in
> this setup.
>
> FE
>
> On Thursday, March 22, 2001 1:45 AM, Greg Matthews
> [SMTP:[EMAIL PROTECTED]] wrote:
> > dear all,
> >
> > is there a way to get the server.xml defined  "name"
> attribute?
> >
> > after following a previous post, i can get the application.xml
> 
> > value but this is not quite what i need.
> >
> > the plan is to be able to deploy multiple copies of the same application
> > for different clients, and have an initServlet read in the application
> name
> > to determine where the config files live.
> >
> > e.g. client1 => read in \\someMachine\config\client1\config.xml
> >client2 => read in \\someMachine\config\client2\config.xml
> >
> > this way, if i have to cluster client1, then there's only ever a single
> > copy of their config information on the network and i don't have to
> > duplicate it per orion installation.
> >
> > i'm also thinking of putting the web-site.xml files for each client
> > in their own config directory to further reduce duplication of
> > config details that would appear if you do clustering.
> >
> > i'd also like to avoid having to alter servlet initialisation parameters
> > or anything in application.xml when deploying the app for a new
> > client, but would consider these as a last resort.
> >
> > thanks,
> > greg.
> >  << File: ATT5.html >>
>

ASP config + clustering

[Greg Matthews]

dear all,
 
is there a way to get the server.xml defined 
 "name" attribute?
 
after following a previous post, i can get the 
application.xml 
value but this is not quite what i 
need.
 
the plan is to be able to deploy multiple copies of 
the same application
for different clients, and have an initServlet read 
in the application name
to determine where the config files 
live.
 
e.g. client1 => read in \\someMachine\config\client1\config.xml
   client2 => 
read in \\someMachine\config\client2\config.xml
 
this way, if i have to cluster client1, then 
there's only ever a single
copy of their config information on the network and 
i don't have to 
duplicate it per orion installation.
 
i'm also thinking of putting the web-site.xml files 
for each client
in their own config directory to further reduce 
duplication of 
config details that would appear if you do 
clustering.
 
i'd also like to avoid having to alter servlet 
initialisation parameters
or anything in application.xml when deploying the 
app for a new
client, but would consider these as a last 
resort.
 
thanks,
greg.

Re: How about the latest xerces.jar and xalan.jar? - memory leak

[Greg Matthews]

Dear all,

Just a warning to anyone using xerces in their web app.

The current version of xerces shipped with orion sort of leaks memory
(xerces issue, not orions).

org.w3c.Document objects continue to hold a reference to any
org.w3c.traversal.TreeWalker objects created so if the TreeWalker goes out
of scope it's not garbage collected.

The latest release of Xerces, Document doesn't hold a reference to
TreeWalker.

In a scenario where:

1. your web app through a servlet or jsp loads a xml config file as a static
variable
2. on subsequent posts or gets creates a TreeWalker to go over the Document

you will be *leaking memory* unless you manually call removeTreeWalker in a
xerces specific way.

If orion shipped the latest xerces/xalan then this problem would go away and
we'd have access to better transformation libraries, etc.

We've tried replacing xerces.jar with the latest version. It fixes the
memory leak problem but xalan falls over

Greg.

- Original Message -
From: "Ted R. Rice" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, March 10, 2001 4:58 AM
Subject: RE: How about the latest xerces.jar and xalan.jar?


> i've experience similar issues with orion 1.4.7 and
> saxon 6.2.1.
>
> if you update the versions of xerces to 1.3 and xalan
> to 2.0, you will get the MethodNotFound error when
> doing any javax.xml package calls. only way i got
> around it was:
>
> 1. remove jaxp.jar from the orion dir
> 2. open up xalan.jar and remove all the javax.xml classes
>
> additionally, we use UTF-8 encoding on all of XML/XSLT
> files to support localization. if you set the default-charset in
> the global-web-applicaiton.xml to be UTF-8 it will not work
> as it complains about unrecognized UTF-8 characters in the
> XML files. hoewever, the second you remove the default-charset
> it works perfectly.
>
> everything i've mentioned was never a problem until 1.4.7. none
> of our code has changed and it never had an issue with the
> default-charset,
> jaxp, or xalan/xerces.
>
> ted rice
>
> -Original Message-
> From: Duffey, Kevin
> Sent: Fri 3/9/2001 9:42 AM
> To: Orion-Interest
> Cc:
> Subject: How about the latest xerces.jar and xalan.jar?
>
>
>
> Orion team (and whoever else is interested),
>
> I am working on a small framework that uses Saxon and requires
> the latest
> xerces.jar file. I just did an autoupdate and it downloaded
> xerces.jar into
> the orion folder (I deleted the one that was there because of
> the problem I
> am about to describe). When I started my app, the MethodNotFound
> error I got
> when I left the 589KB version of xerces.jar in the Orion folder
> came back.
> So two things. First, can you please add the latest xerces
> 1.3.0. Its 1.5MB
> in size, but it adds a lot of new functionality that some of us
> are now
> using. Secondly, why is it when xerces.jar is in the Orion
> folder, Orion
> does NOT use the version I have in the WEB-INF/lib folder? I
> thought (sorry
> if I don't recall what the specs say on this) that libraries in
> the
> WEB-INF/lib of a web-app override the ones in a server dir?
>
> Thanks.
>
>
>
>

Re: SSL Certificates + Security Alert

[Greg Matthews]

thanks to those that helped out.

i checked back with the thawte site and i'd not installed
the root test certificate.

now that the thawte root test certificate is installed i can
go from non-secure to secure page without getting
the security alert.

thanks.
greg.

- Original Message -
From: "Tim Endres" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Monday, March 12, 2001 3:38 PM
Subject: Re: SSL Certificates + Security Alert


> > Q: do i get the dialog window appearing because i'm using a test
certificate? i.e. will the use of a live certificate mean that i don't get
the dialog window appearing?
> >
>
> Yes, *if* the certificate chain of that certificate leads to a root
> certificate that is installed in the user's browser.
>
> tim.
>
>

SSL Certificates + Security Alert

[Greg Matthews]

dear all,
 
i've pretty much got secure pages going with a 
*test* certificate but receive a windows "security alert" dialog box the *first* 
time i hit a page in the secure site.
 
if i click "yes" to "do you want to proceed?" then 
everything seems to work ok and i get the padlock icon in the lower right hand 
corner of the browser, and https://whatever in 
the address bar.
 
Q: do i get the dialog window appearing because i'm 
using a test certificate? i.e. will the use of a live certificate mean that i 
don't get the dialog window appearing?
 
thanks,
greg.

Re: security in web-apps

[Greg Matthews]

there shouldn't be a problem with this.

you might want to post more information on what url you're using + the
config details in server.xml and default-web-site.xml.

- Original Message -
From: "Kiss Tibor" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Saturday, March 10, 2001 5:49 PM
Subject: security in web-apps


> Hi!
>
> I have created an ear file containing only a war file (with some jsps)
> just for experimenting, as I am quite new to orion.
> However, if I try to access my web-app, I get a 403 Forbidden error.
>
> Anyone got some suggestions/examples how I can control access security
> for my war?
>
> Thanks
> Tibor
>

Re: EJB Clustering

[Greg Matthews]

 
william,
 
look for the clustering guide on www.orionserver.com in the 
documentation section,
then maybe post some more questions once you've 
given it a try.
 
it works like you describe, in that starting the 
load balancer takes all requests
and redirects to instances of orion.
 
cheers,
greg.

  - Original Message - 
  From: 
  William Jones 
  To: Orion-Interest 
  Sent: Friday, March 09, 2001 7:32 
PM
  Subject: EJB Clustering
  
  I 
  would like to find out if anybody out there has done much with clustering EJB 
  (or RMI) servers using Orion.
   
  We 
  will be running an operation that will need to be large, scalable and have a 
  failsafe level of availability.  In other words there should be no single 
  point of failure.
   
  Obviously running a single EJB server will lead to a single point of 
  failure in the infrastructure (and obviously limits scalability also).  
  What we sould like to do is have the ability to run several clustered EJB 
  servers so as to provide the ability to scale (add more servers to the 
  cluster) and failover (if one falls the others pick up the slack).  
  
  The 
  classic solution to this sort of clustering (e.g. when done with web servers) 
  is to place a local director in front of the cluster.  Then all requests 
  are sent the the IP of the local director which just routes them through to 
  one of the machines in the cluster.  The obvious problem when doing this 
  with RMI/EJB is that when the remote interfaces (well, stubs really) do a 
  bind() into JNDI the IP address of the machine is bound in with it.  This 
  means that requests made through that stub will go back to the machine doing 
  the registering and not the local director, which defeats the local 
  director.
   
  Weblogic supports clustering by providing its own (proprietry, 
  yuk) version of rmic (c.f. http://www.weblogic.com/docs51/classdocs/API_rmi.html#compiler), but I am not aware of anything similar in Orion.  (By the 
  way, please can no-one interpret this comment as being anything of the "WL is 
  better than Orion" sort... it is clear most of us that Orion is superior in 
  most aspects).
   
  For 
  RMI we have implemented a solution that invloves binding different instances 
  of the same RMI service as different entries in the same JNDI directory and 
  using a central service (or array of them) to pass out the stubs from the JNDI 
  tree in a round robin fashion to clients that request them.  (e.g. two 
  services to implement logging bind to 
  "rmi/com/amc/system/logging/LoggingService/bart" and 
  "rmi/com/amc/system/logging/LoggingService/homer", clients make a request for 
  com.amc.system.logging.LoggingService.class and receive back a stub for one or 
  there of the services).
   
  I 
  can't see that this approach will work with EJBs as you don't seem to have 
  control over to where in the JNDI tree the stubs are bound (the mapping of 
  JNDI name to remote interface is done in the 
  application-client.xml!)
   
  Can 
  anyone give me any insights on other ways to approach clustering of EJB 
  services using Orion?
   
  Thanks in advance...
   
  William Jones
  asset-management.com
   

Fw: Your message concerning Re: non-secure and secure site

[Greg Matthews]

can someone at orion please take a look at removing
an email address, [EMAIL PROTECTED] from the
mailing list.

i don't know about other subscribers but whenever
i forward a message to orion interest, i get a response
from netch.se about some mail account not existing.

apparently they've tried unsubscribing through the
form on www.orionserver.com but it didn't work.

- Original Message -
From: "Stefan Nyström" <[EMAIL PROTECTED]>
To: "Greg Matthews" <[EMAIL PROTECTED]>
Sent: Thursday, March 08, 2001 8:40 AM
Subject: Re: Your message concerning Re: non-secure and secure site


>
> > Date: Thu, 8 Mar 2001 08:27:59 +1000
> > From: Greg Matthews <[EMAIL PROTECTED]>
> >
> > have you tried the following?
> >
> > 1. www.orionserver.com
> > 2. click on "mailing list"
> > 3. enter the email address to unsubscribe, then press the "unsubscribe"
> > button.
> >
> > if so, let me know and i'll contact orion to have the offending email
> > address removed.
>
> Yes I have.  I would appreciate your assistance.
>
> /S.
>
> --
>   _ __  __  _  _  | Stefan Nyström
>  (_  | (__ |__ /_\ |\ | [EMAIL PROTECTED]  ---
>  __) | (__ |  /   \| \| http://www.netch.se/~stefan/(o o)
> -ooO-(_)-Ooo
> "Le concert de finito, finale, you must a go home now.  I have told you so
> many times.  Saa er det dags og gå hjem".  -- Bob Hund
>

Re: non-secure and secure site

[Greg Matthews]

looking at the online documentation, it seems (???) 
like this isn't really to do with SSL.
 
i gave it a try anyway and got the windows dialog 
"save or download" appearing when i tried to open a jsp in the "/secure/" 
directory.
 
does this work? i.e. have to put pages in the 
"/secure/" directory, and when you try to open them it switches to https 
?
 
thanks,
greg.


  - Original Message - 
  From: 
  cybermaster 
  To: Orion-Interest 
  Sent: Wednesday, March 07, 2001 1:14 
  AM
  Subject: RE: non-secure and secure 
  site
  
  
  Greg, 
  all can be in one app. You have to specify a  in 
  web.xml, best specified as a sub-directory e.g. “/secure/*”, where you put 
  your secure pages.
   
      
  --peter
   
   

non-secure and secure site

[Greg Matthews]

dear all,
 
what's the standard way to structure an app where 
some pages are non-secure and other are secure?
 
i can set up a secure site, and redirect from the 
non-secure site to the secure site, and wondered what the best way is to 
transfer context to the secure site to continue processing.
 
e.g.
 
non-secure => search for products
non-secure => add products to cart
non-secure => press checkout, write cart details 
to database
non-secure => redirect to secure-site sending 
cart number
secure  => get 
user to login again
secure  => reload 
cart details from database
secure  => verify 
shipping address
secure  => process 
credit card through payment gateway
secure  => redirect 
back to non-secure site
 
this would also involve duplicating a lot of the 
code between the applications that would run in the secure and non-secure sites, 
like code to read/write cart contents, read/write user details for 
login.
 
is there a better way?
 
are they any pitfalls i've missed?
 
thanks,
greg.

Re: significance of web-site.xml

[Greg Matthews]

good question. clustering islands seem to be only specified in a
web-site.xml file. i'm more of a java programmer than a network/webserver
administrator so apologies if this is a bit of a dumb question.

if i want to run copies of my app for different paying customers, and
cluster each individually then i'm under the impression that i need each
copy of the app configured in separate web-site.xml files.

i understand (and have tried) everything that you've included in your email
but i still don't have an answer to my question.

does creating a new web-site.xml literally correspond to a new registered
domain name?

is there a way to have,

www.ourdomain.com/customer1/myApp
and
www.ourdomain.com/customer2/myApp
and so on

setup, where customer1 and customer2 have been set up in separate
web-site.xml files, and are also clustered in separate islands.

i've got clustering going, and i've got multiple apps configured but all
using default-web-site.xml which is not what we're after.

i've also followed the virtual-hosts tutorial but it also didn't go the
extra step and explain what a new web-site.xml meant.

thanks,
greg.


- Original Message -
From: "Jeff Hubbach" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Monday, March 05, 2001 2:38 PM
Subject: Re: significance of web-site.xml


> Greg,
>
> I guess the first question is why are you trying to do this? The web-site
is associated with a distinct web site. If you have multiple applications
you want under the same web site, ie www.blah.com, then you just make
multiple applications in the web-site file that points its host to
www.blah.com. If you want 2 distinct websites, ie www.foo.com and
www.bar.com, then you make 2 separate web-site.xml files and point the host
of each one to these separate URLs.
>
> Jeff Hubbach
>
> On Mon, 5 Mar 2001 09:22:51 +1000
> "Greg Matthews" <[EMAIL PROTECTED]> wrote:
>
> >dear all,
> >
> >i've been trying to run multiple copies of an application
> >with each copy having its own config\-web-site.xml file.
> >
> >what's the significance of doing this? do i require a network card
> >for each app that i want to run this way?
> >
> >i can run multiple apps in they're all configured in the
default-web-site.xml
> >file but not when each app has it's own web-site.xml file.
> >
> >i believe this has something to do with the "host" attribute within
> >the  element, but don't know what to change it to.
> >
> >i've read the various faqs on orionserver and orionsupport, and
> >read the virtual-hosts faq, but not had any success.
> >
> >thanks,
> >greg.
> >
>

significance of web-site.xml

[Greg Matthews]

dear all,
 
i've been trying to run multiple copies of an 
application
with each copy having its own 
config\-web-site.xml file.
 
what's the significance of doing this? do i require 
a network card
for each app that i want to run this 
way?
 
i can run multiple apps in they're all configured 
in the default-web-site.xml
file but not when each app has it's own 
web-site.xml file.
 
i believe this has something to do with the "host" 
attribute within
the  element, but don't know what 
to change it to.
 
i've read the various faqs on orionserver and 
orionsupport, and
read the virtual-hosts faq, but not had any 
success.
 
thanks,
greg.

Re: multiple apps

[Greg Matthews]

thanks for the response.
 
i'm more of a java programmer than a web site 
administrator so i'm a bit in the dark here.
 
does that mean that i need two network cards in my 
machine if i want to use 2 
appname-web-site.xml files?
 
you suggest changing the host attribute from [ALL], 
what should i change it to?
 
the larger picture is that we're trying to find the 
best way to set up orion so we
can run duplicate copies of an application, one for 
each time a person calls us,
gives us their credit card number, and says "i'd 
like to lease your browser based
software"
 
we want to be able to:
    - run multiple apps per machine, 
and
    - run an app across multiple 
machine where a customer puts heavy load on the system.
 
 
thanks,
greg.

  - Original Message - 
  From: 
  Michael Bosch 
  
  To: Orion-Interest 
  Sent: Wednesday, February 28, 2001 5:04 
  PM
  Subject: Re: multiple apps
  
  Having just gone through learning all the 
  configuration details recently the thing that pops out in my mind is whether 
  or not you changed the host attribute in your web-site element of the 
  copy1-web-site.xml and copy2-web-site.xml files so that they're 
  different?  It'd make sense that one would run and the other doesn't when 
  they're both activated if this were the case.  If they're both set to [ALL] then the first one would probably override 
  the 2nd.  
   
  -Mike
  
- Original Message - 
From: 
Greg 
Matthews 
To: Orion-Interest 
Sent: Tuesday, February 27, 2001 5:27 
PM
Subject: multiple apps

 
Dear all,
 
I'm having some trouble getting multiple copies 
of an app to run where 
i use different -web-site.xml 
files.
 
I don't have a problem where i configure 
all the information within
server.xml and default-web-site.xml, but i want 
to use different
web-site.xml files so that i can do 
clustering.
 
I do the following:
 
1. copy/paste the application within 
\orion\applications.
    e.g.    
\orion\applications\copy1
        
 \orion\applications\copy2
 
2. create 2 web-site.xml files
 
    e.g.    
\orion\config\copy1-web-site.xml
        
 \orion\config\copy2-web-site.xml
 
3. put a reference to the web-site files in 
server.xml
 
  
4. put a reference to the 2 applications in 
server.xml
 

  
 
 
Questions
 
 
a) I can run the copy1 app in my browser, 
but not copy2.
 
b) If i comment out copy1, then for some 
reason i can run copy2.
 
c) If i uncomment copy1, then copy1 starts 
working, and copy2 is not accessible.
 
d) I don't get any error messages relating 
to not being able to access copy1 or copy2.
 
 
Does anyone know what i might be doing 
wrong?
 
Thanks,
Greg
 

multiple apps

[Greg Matthews]

 
Dear all,
 
I'm having some trouble getting multiple copies of 
an app to run where 
i use different -web-site.xml 
files.
 
I don't have a problem where i configure all 
the information within
server.xml and default-web-site.xml, but i want to 
use different
web-site.xml files so that i can do 
clustering.
 
I do the following:
 
1. copy/paste the application within 
\orion\applications.
    e.g.    
\orion\applications\copy1
        
 \orion\applications\copy2
 
2. create 2 web-site.xml files
 
    e.g.    
\orion\config\copy1-web-site.xml
        
 \orion\config\copy2-web-site.xml
 
3. put a reference to the web-site files in 
server.xml
 
  
4. put a reference to the 2 applications in 
server.xml
 

  
 
 
Questions
 
 
a) I can run the copy1 app in my browser, but 
not copy2.
 
b) If i comment out copy1, then for some 
reason i can run copy2.
 
c) If i uncomment copy1, then copy1 starts 
working, and copy2 is not accessible.
 
d) I don't get any error messages relating to 
not being able to access copy1 or copy2.
 
 
Does anyone know what i might be doing 
wrong?
 
Thanks,
Greg
 

Re: load balancing

[Greg Matthews]

found it.
 
someone else was using clustering also and we 
weren't specifying minimumIsland and maximumIsland when launching loadbalancer, 
i.e. each of our loadbalancers were 
seeing each others servers.
 

  - Original Message - 
  From: 
  Greg 
  Matthews 
  To: Orion-Interest 
  Sent: Monday, February 26, 2001 12:52 
  PM
  Subject: load balancing
  
  dear all,
   
  i've had a go at setting up load balancing and 
  it's half working.
   
  i don't know what's registering with the load 
  balancer for the first server mentioned in the message below. the second 
  server mentioned is the one i configured.
   
  anyone have a rough guess about what i might have 
  done to cause the first server to be appearing.
   
  when i open pages on the site, the load balancer 
  finds 127.0.0.1 invalid, then removes it from the cluster, but then about 5 
  seconds later it re-adds it back in.
   
  F:\orion>java -jar -debug 
  loadbalancer.jar
  Balancer initialized...Added island number 1 
  to the cluster...Discovered server 
  127.0.0.1/127.0.0.1:8081...    <= where's this coming 
  from?Discovered server PIII-450/57.226.58.6:8080...    
  <= this is ok.
   
   
  from default-web-site.xml
   
  
      
      
   
      
   
  

load balancing

[Greg Matthews]

dear all,
 
i've had a go at setting up load balancing and it's 
half working.
 
i don't know what's registering with the load 
balancer for the first server mentioned in the message below. the second 
server mentioned is the one i configured.
 
anyone have a rough guess about what i might have 
done to cause the first server to be appearing.
 
when i open pages on the site, the load balancer 
finds 127.0.0.1 invalid, then removes it from the cluster, but then about 5 
seconds later it re-adds it back in.
 
F:\orion>java -jar -debug 
loadbalancer.jar
Balancer initialized...Added island number 1 to 
the cluster...Discovered server 
127.0.0.1/127.0.0.1:8081...    <= where's this coming 
from?Discovered server PIII-450/57.226.58.6:8080...    <= 
this is ok.
 
 
from default-web-site.xml
 

    
    
 
    
 

ejbFindByPrimaryKey

[Greg Matthews]

Dear all,
 
I'm using BMP and noticed that ejbFindByPrimary key 
is called
after ejbLoad.
 
The order in which methods are called are as 
follows:
 
1. ejbFindAll 
2. ejbLoad called once for each entity
3. ejbFindByPrimaryKey called once for each 
entity.
 
I'm pretty sure the container is calling ejbLoad 
(which seems normal if
the entity has never been previously loaded) and 
then ejbFindByPrimaryKey.
 
Why would ejbFindByPrimaryKey be called if the 
container has just called ejbLoad?
 
Thanks,
Greg.

findByPrimaryKey checking db every time

[Greg Matthews]

dear all,
 
is there any way to prevent findByPrimaryKey 
hitting the db every single time it gets called.
 
i.e. if the bean is in cache, can you force the 
container to just return it without
  doing a primary key lookup in the 
db?
 
greg.

1 app connecting to multiple identical db's

[Greg Matthews]

dear all,
 
i have a question about running multiple identical 
database in an ASP (application service provider) environment.
 
we currently use stateless session beans only and 
handle database connection pooling ourselves.
 
this allows us to have a single ejb server 
connection to 1 of several identical databases.
 
new clients have a new database created for 
them, and their config information entered into our own config 
file.
 
based on their login, we run the ejb method using a 
connection from a connection pool pointing at their db.
 
how would we do the equivalent to this if we were 
to rewrite some parts of the system as entity beans?
 
the problem is that i can't see how their login 
information would get propagated down to entity bean level so that i could get a 
reference to the correct datasource. it's easy enough when you're just using 
SLSB's.
 
e.g. pseudo code for what we do 
currently.
 
    public someMethod( Identity id ) 
{
 
        get 
connection from pool based on id.databaseId
        pass the 
connection down as a parameter to each method that requires db 
access
        return 
connection to pool
    }
 
 
 
how would this work with entity beans?
 
 
e.g. dodgy code, but just to show you roughly what 
i mean.
 
private Connection getConnection() throws 
SQLException, RemoteException {
 
DataSource ds = 
null;
 
    // <<>>
    // e.g. int databaseId = 
entityContext.getDBId()
 
    try {    
    Context ic = new InitialContext();    
    ds = (DataSource)ic.lookup( "jdbc/MyAppDS" 
+ databaseId );} catch(NamingException e) 
{        
e.printStackTrace();        throw new 
RemoteException("Could not obtain DataSource: " + e);    
}    return ds.getConnection();}
 
thanks,
greg.

Re: serialisation problem?

[Greg Matthews]

 
thanks juan and mike for the 
responses.
 
F:\dev\script>java -versionjava version 
"1.3.0"Java(TM) 2 Runtime Environment, Standard Edition (build 
1.3.0-C)Java HotSpot(TM) Client VM (build 1.3.0-C, mixed mode)
 
 
is there any way i get see the marshalling code 
generated by orion
so i could maybe work out what's going 
on?
 
if so, can the marshalling code generated by orion 
be modified 
for debugging purposes?
 
thanks,
greg
 

  - Original Message - 
  From: 
  Juan 
  Lorandi (Chile) 
  To: Orion-Interest 
  Sent: Thursday, January 18, 2001 2:32 
  AM
  Subject: RE: serialisation problem?
  
  shooting from the hip again (bad habits die 
  hard)...
   
  String's limit was 64K until JDK 1.3... which JDK are 
  you using?
   
  
-Original Message-From: Greg Matthews 
[mailto:[EMAIL PROTECTED]]Sent: Miércoles, 17 de Enero de 
2001 1:27To: Orion-InterestSubject: serialisation 
problem?
dear all,
 
we're got a strange problem whereby a large 
(141K) String return value from a stateless session bean is turning up at 
the client as an empty string.
 
this happens whether we call the ejb method 
from a jsp or a java client external to orion.
 
it has also happened in other ejb methods 
during testing where a master-detail type record suddenly stopped being 
retrievable when the total size of all data in xml format as a String 
exceeded a certain size.
 
is there an upper limit on the size of 
variables that can be serialised?
 
does anyone have any ideas on why this might be 
happening?
 
thanks,
greg

serialisation problem?

[Greg Matthews]

dear all,
 
we're got a strange problem whereby a large (141K) 
String return value from a stateless session bean is turning up at the client as 
an empty string.
 
this happens whether we call the ejb method from a 
jsp or a java client external to orion.
 
it has also happened in other ejb methods during 
testing where a master-detail type record suddenly stopped being retrievable 
when the total size of all data in xml format as a String exceeded a 
certain size.
 
is there an upper limit on the size of variables 
that can be serialised?
 
does anyone have any ideas on why this might be 
happening?
 
thanks,
greg

NoSuchMethodException

[Greg Matthews]

Dear all,
 
I've build an ejb jar, deployed it, and then tried 
to call one of the methods in one of the stateless session beans.
 
It doesn't work.
 
I can get a reference to the stateless session 
bean but can't work out why i can compile client code to call an ejb method, but 
then get a NoSuchMethodException when i try to run it.
 
 
The line below is from a System.out.println call, 
showing that i've got a remote object reference.
 
testing server = TestingServer stateless 
session  
com.evermind.server.rmi.OrionRemoteException: Invocation error: 
java.lang.NoSuchMethodException    
at 
com.evermind.server.rmi.bd.invokeMethod(JAX)    
at 
com.evermind.server.rmi.a4.invoke(JAX)    
at __Proxy1.initialise(Unknown 
Source)    at 
com.traveltech.data.OrionClient.(OrionClient.java:27)    
at com.traveltech.data.OrionClient.main(OrionClient.java:37)
 
    Nested exception 
is:java.lang.NoSuchMethodException    
at 
com.evermind.server.rmi.bd.invokeMethod(JAX)    
at 
com.evermind.server.rmi.a4.invoke(JAX)    
at __Proxy1.initialise(Unknown 
Source)    at 
com.traveltech.data.OrionClient.(OrionClient.java:27)    
at com.traveltech.data.OrionClient.main(OrionClient.java:37)
 
 
Does anyone have any clues as to why I can't call a 
method in a object that I have a reference to? The method is there since doing a 
javap shows the method, and my client code compiles ok.
 
Thanks,
Greg.

connection pooling

[Greg Matthews]

dear all,
 
do requests for EJB UserTransaction objects or 
db connections accessible through a Context get handled in a *queued* manner 
under orion?, or does it just notifyAll() the worker threads and let them fend 
for themselves?
 
we're currently doing our own database connection 
pooling, about to switch over to UserTransaction objects + using orion's db 
connection pooling.
 
i had to modify the database toolkit we are using 
since it implemented a "free for all" approach to handing out newly freed up 
connections to worker threads, i.e. notifyAll(), which didn't guarantee that 
threads received a db connection in the order in which they were 
requested.
 
greg

Re: Ok. I feel dumb. Data Sources and MS SQL.

[Greg Matthews]

we use a driver from http://www.inetsoftware.de/ for sql server.

there's an evaluation version if you want to give it a go

- Original Message -
From: "Kimberley Scott" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, October 25, 2000 11:15 AM
Subject: Ok. I feel dumb. Data Sources and MS SQL.


> Hiya,
>
> Feeling real stupid here. I'm examining data sources and getting totally
> confused. Maybe I'm just burnt out. Haven't had a holiday in five years.
> Sigh.
>
> Back to business. We're *forced* to use Microsoft SQL 7 and Win2K. How do
I
> set up the datasource xml file to handle this? I can't seem to find
anything
> on this in the email lists on in the docs. The only drivers I have are the
> ODBC bridges (The invisible hidden MS one and the standard Sun one) and
I'm
> not paying for a Type IV driver at the outrageous prices I've been
presented
> with from BEA and the like. Has anyone done this before or can point me in
> the right direction?
>
> A tired developer who's feeling very old and dim...
>
> Kimberley Scott
> Senior Web Developer
> Peakhour Pty Ltd
> http://smartoffice.com.au <<- just built. needs work. ASP/COM Sigh. What
can
> I say?
> http://peakhour.com.au <<- corporate site
> http://www.geocities.com/kimmie_scott <<-me
>
>
>
>

Re: get reference to UserTransaction fails

[Greg Matthews]

i was trying to do the same thing, and it sort of makes
sense that you can't control a transaction (UserTransaction)
from outside orion (i.e. the ejb container).

it does raise the question however about how you would
ever maintain any decent level of control over transactions
if entity beans were involved unless you *always*:

1. set the trans-attribute of every entity bean to Mandatory
2. used stateless session beans as a facade to controlling
   entity beans.

i've been using stateless session beans for a while and trying
to come up to speed with entity beans. there's probably
something stupid i'm doing but i've been unable to find
a solid way to control database transaction boundaries
unless i use the above pattern (slsb call entity beans).

a) does anyone call entity beans directly from a JSP?
or do you call stateless server beans which wrap entity beans?

b) is it possible to control database transaction boundaries
without wrapping EJB methods calls in a UserTransaction.begin()
and UserTransaction.commit() ?

thanks,
greg


- Original Message -
From: "Juan Lorandi (Chile)" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Wednesday, October 25, 2000 3:49 AM
Subject: RE: get reference to UserTransaction fails


> Are your trying to retrieve something from "java:comp" DIRECTLY from code
> not running on Orion?
>
> I have looked (very quickly) at the specs, and I don't think that's
> allowed(seems pretty unsafe to me)
> Perhaps you can build a session bean that preforms the lookup, then passes
> the reference to the client
>
> 
> But I think it's a bad idea anyway... it's kinda like "one and a half"
layer
> programming...
> 
>
> My 2c,
>
> JP
>
> -Original Message-
> From: Klaus Thiele [mailto:[EMAIL PROTECTED]]
> Sent: Martes, 24 de Octubre de 2000 7:10
> To: Orion-Interest
> Subject: get reference to UserTransaction fails
>
>
> Hello,
>
> i'm using transactions within servlets, sessionbeans
> with no problems.
>
> now i'm trying this from an ordinary app-client but it
> caught an exception:
>"java:comp/UserTransaction not found"
>
> what goes wrong?
>
> thanks a lot
>   klaus
>
>
>  import javax.naming.*;
>  import javax.transaction.*;
>  ...
>  TransactionManager tm =
>(TransactionManager)new
> InitialContext().lookup("java:comp/UserTransaction");
>
> --
> Klaus Thiele - Personal & Informatik AG
> mailto:[EMAIL PROTECTED]
>
>  "There's got to be more to life than compile-and-go."
>
>

java:comp/UserTransaction not found

[Greg Matthews]

 
dear all,
 
i'm sort of still learning about entity beans and 
wondered if 
anyone had any pointers for an error i'm 
getting.
 
i receive *no* error when i try to obtain a 
reference to a UserTransaction
from a stateless server bean under 
orion.
 
i receive the following error when i try to obtain 
a reference to a UserTransaction
from a java program running *outside* orion, that 
is similar to CartClient in
the orion examples.
 
javax.naming.NameNotFoundException: 
java:comp/UserTransaction not found
 
 
i can get around this error by having a stateless 
session bean do the UserTransaction
stuff, and act as a facade to the other various entity and session bean methods
but was wondering if there was a way around the 
above problem.
 
thanks,
greg.

Re: findByPrimaryKey() calling store()

[Greg Matthews]

thanks for the response.

i checked again and i'm using BMP (attached is ejb-jar.xml).

i haven't really got my head around how transactions work
with entity beans although doing lots of reading at the moment.

i'm thinking maybe the orion ejb container is treating each
of my entity bean calls as a transaction (as supported by 
profiling the sql server database we're using), and that's
why store is being called (?).

we currently use stateless server beans, and handle 
connection pooling ourselves since all the ejb connection
pooling stuff asn't that mature when we started our project.

we're now relooking at entity beans, etc, and working
out how to do all the simple things like controlling
transaction boundaries, etc, and connection pooling.

greg.

- Original Message - 
From: "James Ho" <[EMAIL PROTECTED]>
To: "Orion-Interest" <[EMAIL PROTECTED]>
Sent: Thursday, October 19, 2000 6:04 AM
Subject: Re: findByPrimaryKey() calling store()


> Hi Greg...
> 
> According to Orion1.4.0's changes.txt it said:
> 
> Optimized CMP entity stores to only store updated fields.
> 
> 
> I am not 100% sure what it means though..maybe it is what u are after...
> 
> James.
> 
> 
> > Greg Matthews wrote:
> > 
> > dear all,
> > 
> > i have a jsp that calls findByPrimaryKey on an entity bean.
> > 
> > the data that the bean represents exists in the database.
> > 
> > it seems like ejbStore is being called as a result of calling
> > findByPrimaryKey.
> > 
> > also, once i have a reference to the bean, calling an accessor
> > method on it also seems to cause ejbStore to be called.
> > 
> > is there something i can set to prevent the bean being
> > updated when there is no (apparent?) need, or am i
> > missing something?
> > 
> > thanks,
> > greg.
> > 
> > // as a result of findByPrimaryKey
> > SELECT id FROM sales_order WHERE id = 5
> > UPDATE sales_order SET notes = 'notes 5' WHERE id = 5
> > IF @@TRANCOUNT>0 COMMIT TRAN BEGIN TRAN
> > 
> > // as a result of calling order.getNotes( ), an EJB method
> > // that does not access the database
> > UPDATE sales_order SET notes = 'notes 5' WHERE id = 5
> > IF @@TRANCOUNT>0 COMMIT TRAN BEGIN TRAN
> 
> 




http://java.sun.com/j2ee/dtds/ejb-jar_1_1.dtd'>


  

  OrderBean
  com.sales.OrderHome
  com.sales.Order
  com.sales.OrderBean
  Bean
  java.lang.Integer
  False
  
	
  reference to a jdbc datasource mentioned down in the DD section


	jdbc/orderDS
	javax.sql.DataSource
	Container
  

  


  

  
	OrderBean
	*
  
  Required

  

findByPrimaryKey() calling store()

[Greg Matthews]

dear all,i have a jsp that calls findByPrimaryKey on an entity 
bean.the data that the bean represents exists in the database.it 
seems like ejbStore is being called as a result of 
callingfindByPrimaryKey.also, once i have a reference to the bean, 
calling an accessormethod on it also seems to cause ejbStore to be 
called.is there something i can set to prevent the bean beingupdated 
when there is no (apparent?) need, or am i 
missing something?
 
thanks,greg.// as a result of findByPrimaryKeySELECT id 
FROM sales_order WHERE id = 5UPDATE sales_order SET notes = 'notes 5' WHERE 
id = 5IF @@TRANCOUNT>0 COMMIT TRAN BEGIN TRAN// as a result of 
calling order.getNotes( ), an EJB method// that does not access the 
databaseUPDATE sales_order SET notes = 'notes 5' WHERE id = 5IF 
@@TRANCOUNT>0 COMMIT TRAN BEGIN TRAN

userThreads + calling EJBs from outside orion.

[Greg Matthews]

 
dear all,
 
i've started orion 1.40 as follows:
 
java -jar orion.jar -userThreads
 
and then tried to get a reference 
to the InitialContext, and received
this stack trace.
 
f:\dev\script>java -classpath 
f:\orion\orion.jar;f:\jdk1.3\jre\lib\rt.jar;f:\classes 
test.OrionTestjavax.naming.NamingException: Not in an application scope - 
start Orion with the -userThreads switch if using user-created 
threads    at 
com.evermind.server.g6.bi(JAX)    at 
com.evermind.naming.j_.lookup(JAX)    
at 
javax.naming.InitialContext.lookup(InitialContext.java:350)    
at 
test.OrionTest.(OrionTest.java:18)    
at test.OrionTest.main(OrionTest.java:31)
 
i'm trying to create a dos shell program that i can 
use to test EJB's
running inside orion.
 
1. am i starting orion correctly to allow this with 
the userThreads flag?
 
2. what values should i use to initialise the 
InitialContext? is the following correct?
 
 Hashtable env = new 
Hashtable(); env.put(javax.naming.Context.INITIAL_CONTEXT_FACTORY, 
"com.evermind.server.ApplicationInitialContextFactory"); env.put(javax.naming.Context.URL_PKG_PREFIXES, 
"com.evermind.server"); Context ctx = new 
InitialContext(env);
 
i know i shouldn't need to if my classpath is set 
up correctly but i'd like to know what values
should be used anyway.
 
thanks,
greg

calling EJBs from outside orion

[Greg Matthews]

dear all,
 
is it possible to call EJBs running in orion from a 
java program started at the DOS command line, i.e from outside 
orion?
 
i guess i might have to generate the client stubs 
(?). has anyone done this?
 
this would be useful when building the ejb 
components to be used by the web developers.
 
thanks,
greg

Re: my orion freezes at a high load

[Greg Matthews]

what database driver are you using?

if you're using the sun jdbc-odbc bridge then it's
prone to locking up and you'll have to purchase a 
proper driver.

-Original Message-
From: Thomas Krantz <[EMAIL PROTECTED]>
To: Orion-Interest <[EMAIL PROTECTED]>
Date: Saturday, 14 October 2000 2:28
Subject: my orion freezes at a high load


>
>Hi,
>
>I am stress testing Orion, and in my application I have
>lots of JDBC calls that use a configured data-source.
>I have noticed that Orion stops serving requests (not even
>static pages) after a while of high load. It still listen
>to port 80, but doesn't return any content. If I increase
>the "max-connections" data-source setting the system lives a little
>bit longer. Now, I realize that I probably have leaking connections,
>but shouldn't Orion log somewhere if the max-connection limit is
>reached?  And why doesn't it serve static pages? I can't find anything
>strange in any logs at all.
>
>I am using Solaris 8 sparc with Java 1.2 (tried 1.3 also) and
>Orion 1.3.7.
>
>Any clues?
>
>/ Thomas
>
>--
>"Hard work often pays off after time, 
> But laziness always pays off now."
>
>PGP Public Key: http://www.algonet.se/~tok/pgp.txt
>
>