RE: Problems specifying persistence directory in 1.5.2
Hi, Did you put that into the bugzilla database? If not, please do so. This seems to be really a bug which prevents you from starting multiple orion server from the same machine when you are using the same binaries directory. Greetings from Munich, /Thomas -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of [EMAIL PROTECTED] Sent: Thursday, June 21, 2001 7:31 PM To: Orion-Interest Subject: Problems specifying persistence directory in 1.5.2 In moving from 1.4.5 to 1.5.2, we are no longer able to successfully specify directory location for application state persistence (see error below). Version 1.5.2 looks for a directory named persistence in the application directory, ignoring any path attribute value, relative or absolute (see below). We tried this on both linux and 2000 with identical results. Any input/confirmation of this problem would be most appreciated. Orion Error: Auto-deploying dio (New server version detected)... Fatal Error: Transaction log file (/D:/development/dio/orion/persistence/transaction.state) did not exist and was not possible to create, the most common reason for this is an invalid path or Orion lacking security to write to that path: /D:/development/dio/orion/persistence/transaction.state (The system cannot find the path specified) orion-application persistence path=./mypersistence/ /orion-application Sam
AW: clustering and key generation
Hello. java.naming.factory.initial=com.evermind.server.rmi.RMIInitialContextFactory java.naming.security.principal=principal java.naming.security.credentials=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? I had the same problem from inside an JSP page, and solved it by adding dedicated.connection=true to the properties. However, I'm not sure if this works from inside an EJB also. (see orion bugzilla bug #176 for this) /Thomas -- hybris GmbH | /Thomas Hertz Frankfurter Ring 162 | Software Engineering D-80807 München | | t +49(0)89 89065-0| [EMAIL PROTECTED] f +49(0)89 89065-555 | www.hybris.de PGP: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E
AW: http-ormi tunneling
Hello. When we try to connect to the server using http:ormi://localhost/appName, the tunnel servlet is initialized (message in $orion-home$/log/global-application.log) but the server gives no response. The pure ormi (ormi://localhost/appName ) connection is working fine. Strrrange, I followed exactly the same path as you did and it worked like a breeze. We have the same problems here. Servlet is loaded, but the client always receives a Error communicating with server: Lookup error: java.net.ConnectException: Connection refused: connect. from the server. This is the same error you will receive if orion is not running. A Firewall cannot be the problem, I tried that with client/server on the same computer. If somebody else had these problems and solved them, please tell us.. Greetings from Munich, /Thomas -- hybris GmbH | /Thomas Hertz Frankfurter Ring 162 | Software Engineering D-80807 Munchen | | t +49(0)89 89065-0| [EMAIL PROTECTED] f +49(0)89 89065-555 | www.hybris.de PGP: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E
AW: http-ormi tunneling
Hi, Maybe something todo with the Orion version? I'm using 1.4.5 on Win98 and WinNT and have to say: it works. Neither 1.4.4 nor 1.4.5/1.4.7/1.48 on WinNT or 2000 works. I'll try a fresh installation. Let's see if this is working... Greetings, /Thomas -- hybris GmbH | /Thomas Hertz Frankfurter Ring 162 | Software Engineering D-80807 Munchen | | t +49(0)89 89065-0| [EMAIL PROTECTED] f +49(0)89 89065-555 | www.hybris.de PGP: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E
error (500) on www.orionserver.com
Hi, if I open the "FAQ" section on www.orionserver.com, I get a 500 Internal Server Error com.evermind.server.rmi.OrionRemoteException: Database error: The database is already in use by another process at KeywordHome_EntityHomeWrapper22.findAll(KeywordHome_EntityHomeWrapper22.java:642) at com.evermind.faq.ejb.EntryManagerEJB.getKeywords(EntryManagerEJB.java:58) at EntryManager_StatefulSessionBeanWrapper3.getKeywords(EntryManager_StatefulSessionBeanWrapper3.java:304) at /viewFAQ.jsp._jspService(/viewFAQ.jsp.java:70) (JSP page line 18) at com.orionserver.http.OrionHttpJspPage.service(JAX) at com.evermind.server.http.HttpApplication.xk(JAX) at com.evermind.server.http.JSPServlet.service(JAX) at com.evermind.server.http.d4.sx(JAX) at com.evermind.server.http.d4.sv(JAX) at com.evermind.server.http.eg.s2(JAX) at com.evermind.server.http.eg.dp(JAX) at com.evermind.util.f.run(JAX) Nested exception is: java.sql.SQLException: The database is already in use by another process at org.hsql.Trace.getError(Trace.java:124) at org.hsql.Trace.getError(Trace.java:115) at org.hsql.Trace.error(Trace.java:127) at org.hsql.Log.open(Log.java:115) at org.hsql.Database.init(Database.java:44) at org.hsql.jdbcConnection.openStandalone(jdbcConnection.java:651) at org.hsql.jdbcConnection.init(jdbcConnection.java:524) at org.hsql.jdbcDriver.connect(jdbcDriver.java:78) at com.evermind.sql.DriverManagerDataSource.getConnection(JAX) at com.evermind.sql.DriverManagerConnectionPoolDataSource.getPooledConnection(JAX) at com.evermind.sql.OrionPooledDataSource.d_(JAX) at com.evermind.sql.aj.d_(JAX) at com.evermind.sql.OrionPooledDataSource.getConnection(JAX) at com.evermind.sql.DriverManagerXADataSource.ev(JAX) at com.evermind.sql.am.er(JAX) at com.evermind.sql.aq.prepareStatement(JAX) at com.evermind.server.ejb.DataSourceConnection.getCustomStatement(JAX) at KeywordHome_EntityHomeWrapper22.findAll(KeywordHome_EntityHomeWrapper22.java:533) at com.evermind.faq.ejb.EntryManagerEJB.getKeywords(EntryManagerEJB.java:58) at EntryManager_StatefulSessionBeanWrapper3.getKeywords(EntryManager_StatefulSessionBeanWrapper3.java:304) at /viewFAQ.jsp._jspService(/viewFAQ.jsp.java:70) (JSP page line 18) at com.orionserver.http.OrionHttpJspPage.service(JAX) at com.evermind.server.http.HttpApplication.xk(JAX) at com.evermind.server.http.JSPServlet.service(JAX) at com.evermind.server.http.d4.sx(JAX) at com.evermind.server.http.d4.sv(JAX) at com.evermind.server.http.eg.s2(JAX) at com.evermind.server.http.eg.dp(JAX) at com.evermind.util.f.run(JAX) Good to see that they are using EJB's;) Greetings, /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
bug: InitialContext(Map) to different servers failed.
Hello, If I try to connect with 'new InitialContext( Map )' to a server and then try to connect to a different server (with a different java.naming.provider.url), the connection is established, but always to the first server I connected to. I did not found a way to connect to more than one server, even if I use the InitialContext.close() before trying to make the new connection. It seems to me that orion caches some Context properties. The orionconsole is able to handle connection to different servers, so I think they are using not 'new InitialContext(Map)' ..? I've posted this to bugzilla. Is this a known bug? Or is it 'by design'? Is there a different way to connect to different servers? Greetings, /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Re: Factoring EJB Homes from within Session EJB
Hello Christian, CB I noticed that I needed to reference all the homes of my EJBs within the CB web.xml and I wanted to know if it was recommended/safe to create one CB stateless session bean which has all the ejb-ref needed in its ejb-jar.xml CB and only reference that stateless session bean from my web app. This way my CB web app only needs to know about one Session Bean and that bean's CB responsibility is factoring other home. CB Does anyone see any problem with this approach? We did that for our project and it works fine. One 'SystemBean' session beans has all ejb-refs, and all other beans only need to know about the SystemBean and can call a getHome( String jndiName ) to obtain home interfaces.. I don't see any disadvantages except of the small performance decrease, because you have an additional bean lookup/method call. But as far as I can see, this decrease is very very small. Greetings from Munich, /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Re: EJB 2.0 with dependent 1 -- n sample available
Hi All, JA As promised, the EJB 2.0 CMP example with 1 to many EB to dependent JA relationships is available. Its posted at: Thanks for this good example, Jim! I have a question: I played with DOs and orion a few months ago, and had some problems: How do I remove DO's? Simply DO.remove() them? I had major problems with that and I am not sure I if was my fault. Does orion support removal of DOs? Does orion support the cascade-delete/ tag in ejb-relationship-role of a DO? This would be very nice. I think I start playing with that and try to extend Jim's example if it is possible. Do you have experience with that? Has somebody examples for this removal process of DO's in Orion? Greetings from Munich, /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
1.4.5. Welcome back!
Thanks, guys. /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Inner classes in JSP don't work
Hi, some weeks ago i found a message in this newsgroup that the bug with inner classes in JSP is now fixed. I am currently using 1.4.4 and it seems that this bug is either not yet fixed or it is back again. I tried also with 1.3.8 and 1.4.4, in all versions I was not able to get a inner class in a jsp file working. Has anybody got the same problem? Should I post it to bugzilla? I am not sure, because Magnus said it was a bug in 1.0rc1 but it fixed now. /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Re[2]: 1.4.4 OR Collection classes
Hallo Robert, RK did you file this as a bug report in bugzilla? I didn't find it there. I'll RK file it now (we just ran into it and it's a nasty one to find). if you RK already have please contact me so I can remove it. RK robert Vidur, This looks in same ways similar to the problem that I reported a couple of days ago. That is, Orion 1.3.8 allowed non-serializable objects to be returned to the client if running in the same JVM (which should of course not be allowed). 1.4.3 on the other hand behaves differently - rather than throwing a java.io.NonSerializableException (?) or something like that (which I suppose would be correct behavour although I haven't looked in the spec), it just returns a null and says nothing. Ein weiteres hartes Problem, das hiermit zusammenhaengt, ist, dass ab der 1.4.4 es vorkommen kann, das ein findXY() NULL zurueckliefert, naemlich genau dann, wenn waehrend der Zeit fuer das zusammenstellen der Collection der Remoteobjekte ein anderer Thread eines der Objekte geloescht hat. Dieses Problem ist nachvollziehbar und in der 1.3.8 noch nicht auf- getreten. Da dieses Szenario (ein Objekt wird geloescht waehrend irgendjemand die Liste der Objekte haben will) durchaus kein seltenes ist, sehe ich dies als schweren Bug an. Angenommen es gibt 1000 entity beans, das findAll() liefert eine Collection von 1000 objekten. falls nun nach dem aufruf von findAll (welches ja eine gewisse zeit dauert) eines der 1000 geloescht wird, sollte doch nicht ploetzlich NULL zurueckgeliefert werden? Ich erwarte entweder 1) eine Collection mit 999 gueltigen objekten (falls das loeschen noch rechtzeitig passierte) oder 2) eine Collection mit 1000 objekten, von denen eines "no longer existent" ist. Wie siehst du das? Ich werde dies mit einem kleinen testcase ins bugzilla stellen, und auch eine Mail in die newsgroup schreiben. /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
1.4.4: findXY() returns NULL if one of the enties is removed
Hello. i filed the following bug to bugzilla: --- The following problem occurs in 1.4.4 but not in 1.3.8: 1) one thread calls a findXY() (lets say findAll() ) on a home interface 2) this will take a while. while thread 1 is processing the findAll() another thread calls remove() for one of the remotes that are expected to be returned by findAll(). 3) the findXY() returnes NULL. in summary: everytime one of the underlaying entities are removed(), the finder methods of orion 1.4.4 always returns NULL. i guess this must be a bug. Robert kruger reported in bug #181 that orion1.4.4 does not allow to return a non serializable value to a client in the same VM. I guess the method which creates the collection for findAll() runs into trouble if an underlaying bean no longer exists, and so the whole process of serialization fails, resulting in a NULL returned to the client. if this problem is not yet known, i can attach a test case. --- what do you think? is this a bug? /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Re: 1.4.4 OR Collection classes
Hi guys, sorry for this mail in german, wrong recipient! It should go to Robert directly. (Hi robert!) for all how are not able to read german, I added a english translation as seperate mail..;) Greetings, /Thomas TH Ein weiteres hartes Problem, das hiermit zusammenhaengt, ist, TH dass ab der 1.4.4 es vorkommen kann, das ein findXY() NULL zurueckliefert, TH naemlich genau dann, wenn waehrend der Zeit fuer das zusammenstellen der TH Collection der Remoteobjekte ein anderer Thread eines der Objekte TH geloescht hat. TH TH [...] and so on and so on german stuff... TH /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Re: DeadlockException
Hello, AC Does anyone know what the following error means? AC 11/14/00 10:51 PM Error in bean com.netregistry.aucom.accounting.BillingAccount: AC Error in ejbPassivate AC com.evermind.server.DeadlockException: Deadlock detected: thread 11 is waiting for AC resource entity 1003 held by thread 11 AC Any ideas on how to track it down? Just ten minutes before I read this message I encountered the same error in my application. (1.4.4) com.evermind.server.rmi.OrionRemoteException: Transaction was rolled back: com.evermind.server.DeadlockException: thread 6 is waiting for resource entity K-5 held by thread 5 in transaction [Transaction 4d:1:0:0:0:0:0:0:0:e2:0:e2:d4:67:89:da] who is waiting for resource entity E-6 held by thread 6 in transaction [ Transaction 4a:1:0:0:0:0:0:0:0:e2:0:e2:d4:67:89:93]; nested exception is: I am not quite sure what that means exactly. Can anybody help us? Thanks in advance, /Thomas -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
uninstalling application
Hello. Is there a possibility to uninstall a application on a remote server with a batch process? I use java -jar admin.jar ormi://host user pwd -deploy for deploying, but haven't found somethink like that for uninstalling. Greetings from munich, /Thomas Hertz -- | /Thomas Hertz | [y] hybris GmbH | Software Engineering | | [EMAIL PROTECTED] | Schwere-Reiter-Strasse 35 | tel +49 89 306697- 0 | | www.hybris.de | Haus 16, D-80797 Muenchen | fax +49 89 306697-99 | | pgp fingerprint: D070 5D86 BE2D C3AF E2CC D2D8 C29A 7F68 7407 629E |
Serializing java.lang.Class Objects
Hello. I'm not sure this is a orion specific issue. Is it true, that serializing a java.lang.Class object and recreating it in a different VM can cause problems? If have a bean which takes a Class[] as arguement. Class implements Serializable, thats not the problem, but when I call the bean, the EJBcode at serverside always receives a NULL-pointer, even when I put something different as an argument. I think that because Class is defined not only by name, but also by it's ClassLoader, it could cause problems when deserializing it? what do you think? How do you transport Class-Object over RMI? Greetings, Thomas - --- Thomas Hertz [EMAIL PROTECTED] hybris GmbH Schwere-Reiter-Str. 35 Gebaeude 16 80797 Muenchen fon: +49 89 30 66 97 0 fax: +49 89 30 66 97 99 http://www.hybris.de solutions for an accelerating generation. -BEGIN PGP SIGNATURE- Version: PGPfreeware 6.5.3 for non-commercial use http://www.pgp.com iQA/AwUBOQVAcMKaf2h0B2KeEQL9OwCeJqxvD179MAwVOQnBpE6od2nM3HEAn2/8 08hHiVxaPb/1HWUyVZWL0vtT =O1ie -END PGP SIGNATURE-
AW: change Transaction Timeout?
Hi All. Well, does orion support user (client) dermarcated transactions at all? I saw this question was already posted and not answered, so I ask again. Does anybody use a session bean with transaction-typeBean/transaction-type For me it looks like orion treats such a bean still as container managed and returnes a java.lang.IllegalStateException: Only beans with user-managed transactio ns can invoke getUserTransaction() if I try to use sessionContext.getUserTransaction(). Any suggestions? Greetings from Munich, thomas Is there a possibility to change the timeout of transactions with orion? I have a method that runs a long time, and I always get a "RMIException: Transaction rolled back: timeout" exception. I guess I can use a UserTransaction and change via setTransactionTimeout(int seconds) this value, but is there a better way to change the xatimeout for methods with container managed transactions? Greetings from Munich, Thomas --- Thomas Hertz [EMAIL PROTECTED] hybris GmbH Schwere-Reiter-Str. 35 Gebaeude 16 80797 Muenchen fon: +49 89 30 66 97 0 fax: +49 89 30 66 97 99 http://www.hybris.de solutions for an accelerating generation.
still Bug in 0.9.4h / 0.9.4l ?
Hello all. A week ago I posted a question to this newsgroup. I think I found a bug in the ejb-module in version 0.94h+. Thanks for your help in advance, Thomas If you need more information, let me know. This is my original message: Hi! I wrote a small application using orion, everything is working fine. I can create/delete entity beans (backed up by a HSQL database) as much as I want. (orion.jar 0.9.4e) But when I try to use the version 0.94h or 0.9.4l everything is different. I am no longer able to create a single ejb. Everytime I try to call a EJBHome.create() (e.g.: ProductCategory cat = productCategoryHome.create( pk ); ) or something like that I get the following on serverside: java.lang.NullPointerException at de.hybris.jakarta.ejb.ProductCategoryEJB.createProductCategory (ProductCategoryEJB.java:65) at EntityBeanWrapper64.createProductCategory(EntityBeanWrapper64. java:790) at java.lang.reflect.Method.invoke(Native Method) at com.evermind.server.rmi.a1.fn(JAX) at com.evermind.server.rmi.a1.run(JAX) at com.evermind.util.b.run(JAX) With the same configuration, with the same ear-file. Whenever I switch back to 0.9.4e or 0.9.4 everything works fine, if I use 0.9.4h or 0.9.4l, the described error occured. The corresponding ejbCreate() method is never called, so the error is somewhere between calling EJBHome.create(params) and EntityBean.ejbCreate(params). What parameters I use doesn't matter, I am not able to create a single EntityBean. Any ideas? Greetings from Munich, Thomas
nullpointer when database connection is lost
Hello. I'm not sure this is a bug, but I don't think an applicationserver should throw a NullPointerException in this case: (Orion/0.9.4) I set up a small entity bean example with HypersonicSQL. Just for testing, I wanted to examine what happens, if orion can no longer access the database . So I closed the connection (stopped the HSQLwebserver) after Orion successfully started. Now every call of a EntityHome.create() method fails with a java.lang.NullPointerException at com.evermind.server.ejb.DataSourceConnection.close(JAX) at EntityHomeWrapper6.create(EntityHomeWrapper6.java:480) at de.hybris.jakarta.ejb.ProductCategoryEJB.createProductCategory(ProductCategoryEJB.java:55) at EntityBeanWrapper5.createProductCategory(EntityBeanWrapper5.java:485) at java.lang.reflect.Method.invoke(Native Method) at com.evermind.server.rmi.a2.fq(JAX) at com.evermind.server.rmi.a2.run(JAX) at com.evermind.util.b.run(JAX) The code fragment of ProductCategoryEJB.java:55 where I call the underlaying 'create' is ... ProductCategory cat = productCategoryHome.create( id ); ... If the HSQL-Server runs everything works fine, if I close it, I get the NullPointerException. When I wait until the ejbPassivate() is called on the enterprise java bean, I get a normal "Connection to server refused: No further information" exception. If you have any idea or need more information, please let me know. Greetings from Munich, Thomas Hert --- Thomas Hertz [EMAIL PROTECTED] hybris GmbH Schwere-Reiter-Str. 35 Gebaude 16 80797 Muenchen fon: +49 89 30 66 97 0 fax: +49 89 30 66 97 99 http://www.hybris.de solutions for an accelerating generation.
