WD Java RFC metadata cache invalidation without recycle JAVA

Since we are in a high availability production system, we cannot recycle or CE instance. There are times were the RFC metdata invalidation does not work and the only things that forces it is a recycle of the CE system.
This is contrary to what SAP has publicized, so is there a process in NWA or a single process (and NOT the server0 process) that we can refresh without stopping and starting the java engine.
Much like Visual Administration is there a single service we can stop and start to ensure the engine stays up and running and available?
Thanks
Weyland Yutani

Hi Weyland ,
did you check the guide ?
Metadata Cache Invalidation for Adaptive RFC Models
http://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/10465350-b4f5-2910-61ba-a58282b3b6df
There's a solution in this guide which explains how to proceed to avoid engine restart in case of ARFC.
(I don't know your engine release, but if it is >= 711, then you may welcome this guide as well:
Using an Adaptive RFC 2 Model in Web Dynpro Java
http://www.sdn.sap.com/irj/scn/index?rid=/library/uuid/50f7192d-2808-2d10-189b-df3fa5f71abf )
I hope this helps.
Best Regards,
Ervin

Similar Messages

Metadata Cache Invalidation for Adaptive RFC Models

Hi,
Im trying the tutorial of the document: Metadata Cache Invalidation for Adaptive RFC Models.
Ive downloaded and deployed the EAR for NW7.0 on my server, but when I try to list the JCO destinations I have the following error:
java.lang.NoSuchFieldError: flush_cache
at com.sap.invalidation.InvalidationView.onActionGetSLDJCOs(InvalidationView.java:215)
When I try to invalidate a dictionary, I have the message:
Cache for Dictionary com.unilog.product.search.model.types.SearchProductModel invalidated
But the tooltip value is not updated in my application...
Can anyone give me some tips to use this component?
Thanks in advance.
Regards,
Claire

hi friends
u see this weblogs&pdf,this is helpful for u
/people/bertram.ganz/blog/2007/01/08/the-top-12-hottest-sap-notes-every-web-dynpro-java-developer-must-know
https://www.sdn.sap.com/irj/servlet/prt/portal/prtroot/docs/library/uuid/f3f93ee7-0c01-0010-2593-d7c28b5377c2
Regards'
vino

Invalidating rfc metadata cache

hi all,
i want to invalidate rfc metada cache through adaptive rfc metada cache invalidation tool in webdynpro console..
can any one plz suggest me the path to find that.
thanks in advane
seshagiri

hi juan,
it was very helpful solution to our problem that is invalidating rfc metadata.
after i gone through the pdf,i found it is for NW04S or 7.0 SP12
but v r using SP09
so i beleive upgraing is the only solution.right
or can u suggest any other solution.
thanks in advance
seshagiri

How to install the metadata cache invalidation tool

Hi,
I have a problem in using JCO destinations for Adaptive RFC models.
I went through the below pdf and they mentioned metadata cahce invalidation tool and it is not installed in the portal
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/10465350-b4f5-2910-61ba-a58282b3b6df
please let me know how to install metadata cache invalidation tool
points will be rewarded for sure for the helpful answers

I installed the Meta data cache invalidation tool using SDM and when i am trying to run that application from Content administrator it is throwing an exception
The currently executed application, or one of the components it depends on, has been compiled against class file versions that are different from the ones that are available at runtime.
If the exception message indicates, that the modified class is part of the Web Dynpro Runtime (package com.sap.tc.webdynpro.) then the running Web Dynpro Runtime is of a version that is not compatible with the Web Dynpro Designtime (Developer Studio or Component Build Server) which has been used to build + compile the application.*
Note: the above hints are only a guess. They are automatically derived from the exception that occurred and therefore can't be guaranteed to address the original problem in all cases.
please let me know how to resolve this issue
Points will be awarded for sure
Bala

How to install metadata cache invalidation tool

Hi,
I have a problem in using JCO destinations for Adaptive RFC models.
I went through the below pdf and they mentioned metadata cahce invalidation tool and it is not installed in the portal
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/10465350-b4f5-2910-61ba-a58282b3b6df
please let me know how to install metadata cache invalidation tool
points will be rewarded for sure for the helpful answers

This tool is already installed on your system.
1) navigate to http://<hostname>:<port>/index.html
2) Click on "Webdynpro"
3) Click on "Web Dynpro Console"
3) Login as an Administrator (any user with admin priv's is ok).
4) Notice the last entry in the list

Deployment: Metadata Cache Invalidation for Adaptive RFC Models

Dear SDN Member,
I have deployed the new "Cache Invalidation" Tool at our portal NW04s SPS12.
We have deployed the tool by using the sdm.
https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/10465350-b4f5-2910-61ba-a58282b3b6df
The entrypoint is not visible in the webdynpro console navigation.
What have I done wrong?
Best regards,
Olaf Reiss

Hello,
we also deployed it to our NW04s SPS12 Portal (since it was NOT included in SPS12)
We deplyoed the correct version. For us it is also not visible in the NWA.
However, it can be called from the Webdynpro console. Then you see it has this direct link.
http://<server>:<port>/webdynpro/dispatcher/local/InvalidationTool/Invalidation
However, the thing still has bugs.
Clicking the button "Get JCO destinations" just crashes with a 500 error (see below). Our developer said
<i>The initial exception that caused the request to fail, was:
   java.lang.NoSuchFieldError: flush_cache
    at com.sap.invalidation.InvalidationView.onActionGetSLDJCOs(InvalidationView.java:215)
    at com.sap.invalidation.wdp.InternalInvalidationView.wdInvokeEventHandler(InternalInvalidationView.java:236)
    at com.sap.tc.webdynpro.progmodel.generation.DelegatingView.invokeEventHandler(DelegatingView.java:87)
    at com.sap.tc.webdynpro.progmodel.controller.Action.fire(Action.java:67)
    at com.sap.tc.webdynpro.clientserver.window.WindowPhaseModel.doHandleActionEvent(WindowPhaseModel.java:420)
    ... 26 more</i>
Using the lower part and option 2.
("Dictionaries Cache Invalidation" and "Enter Dictionary") is said to be working, if you exactly know the naming.
But how can we invalidate the RFC caches used by Visual Composer?
In Visual Admin/services/Connector Container/../SapFactory/Managed conenction factory/ConnectionDefintion I have already set the Connectionlifetime to 60 and ticked "expiraation" ?
Best regards
Ulf

J2EE RFC Metadata cache refresh

Hi All,
I've searched a bit on this forum but can't quite find the answer I'm looking for...
Is it possible to perform a cache refresh without having to restart the server, after making changes to an RFC model and deploying it? It's a pain in the neck to have to keep restarting DEV, QA and PROD everytime we move something through the CMS that has changes to models.
I would have thought the assembly step from consolidation to production would have solved this problem but it doesn't.
Thanks,
Gareth.

> Hi Gareth,
> Several stacks like Web Dynpro, Metadata Interface,
> e, JCO etc read the ABAP metadata and cache it on the
> J2EE Engine. Hence the older metadata still remains
> cached.
> Solution:
> 1) If the interface of a function module is changed
> on the ABAP side, the restart of the J2EE application
> server updates the meta data caches on Java side. A
> restart of the ABAP application server is not
> necessary.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> 2) People think about a option to update the meta
> data cache without a restart of the J2EE application
> server. But nothing in detail is planned at the
> moment.
>
>
>
>
>
>
> Regards,
> Arun
Hi Arun,
I understand what is happening and how it works and why... I was simply asking if there is a way to refresh the J2EE cache without restarting the server?
I think this is something we really need included in future SP's as it is a pain. As mentioned, I would have throught the assembly/import step in the NWDI CMS would have done something to refresh the cache when an SC is imported into the Production system.
Gareth.

Invalidate ARFC Metadata Cache

Hi,
I am working on NW 7.0 S 12.
Everytime I make changes to my Adaptive RFC I require to take a J2EE server restart to invalidate the metadata cache. I am aware of the 'Invalidation of ARFC Metadata Cache link under Web Dynpro Console on the Porta
I am unable to find this link under Web Dynpro Console -> Admin Menu.
I found the document [Metadata Cache Invalidation for Adaptive RFC Models|http://www.sdn.sap.com/irj/scn/index?rid=/library/uuid/10465350-b4f5-2910-61ba-a58282b3b6df]
on SDN by Arun Bhat J which says I can download EARs from SDN for NW7.0 and deploy as a Web Dynpro application.
I have tried this activity by importing the file structure in NWDS as a Web Dynpro Application/ J2EE Application/ Java Application. None of these helped me deploy file structure.
I would like to make this link of 'Invalidation of Cache' available on the portal. How can I achieve this. Also, I have downloaded the files mentioned in the document. How can I use this to solve my problem.
Thanks in advance.

Leo van Hengel wrote:
> However my real advice would be to patch your portal. SP12 (2007) is really old
>
> Cheers,
> Leo
We have the same problem in CE7.2.
Everytime when we change structure in ABAP or in WS - we have to restart CE
cliaring JCO Cache in nwa "JCO monitoring" helps 50/50...
Edited by: Pavel Korgenevskiy on Feb 3, 2012 6:01 AM

Adaptive RFC... metadata cache.

Just curious...Is there a way to force a re cache of the RFC metadata.
Regards
Pran

Hi Pran,
currently, the only way to "recache" or initialize the Metadata cache is by restarting the Netweaver J2EE Engine.
As a matter of fact, whenever (structural) changes are done in a backend (such as adding fields to DDIC Structures), then the engine must be restarted. It is not necessary to reimport, rebuild, or redeploy, but it is necessary to restart
This will be improved in future releases.
Ciao, Markus

Invalidate the Adaptive RFC Dictionary Cache

Hi All,
I had to do RFC interface structure changes for a Webdynpro JAVA application.I had deployed the application and used the administrative function "Invalidate Adaptive RFC Dictionary Cache" within the Web Dynpro console" to explicitly invalidate the metadata cache..
Before using this tool, I was not able to run the application.This tool helped to run the application.
But When I Submit the data to the RFC from this application I get the following Exception:
com.sap.mw.jco.JCO$MetaData.getType(JCO.java:10210)
I understand from the SAP note 859950, Metadata is not invalidated without J2EE Engine restart.
Is there any update from SAP after this note..Pls let me know if some one has done a workaround instead of restarting the SAP NW Application server.
Thanks, Sethu

We went through this same issue many times and we opened an issue with SAP a few months ago.
The invalidate is very limited and does not handle a clustered server architecture. It is meant only for a single node, single server setup. All of our environments are multi-node and multi-server.
The tool will only invalidate the node that the Basis Administrator is currently logged onto. Therefore if your server architecture consists of > 1 node and/or > 1 servers then you will have nodes/servers that still have the previous version of the metadata cached.
SAP said the Basis administrator would have to log onto each node and run the invalidate tool. In a multi-node setup there is no easy way to do this because there is no control which node you will be put on. If you have 1 node per server then yes you can log onto each server and run the tool.
In the end SAP said they do not support the Invalidate tool in a multi-node/multi-server cluster.
They said to perform a J2EE restart, period.
Hope that helps.
Dean Cyril Wood

Error in deploying p6.ear on WebLogic: java.util.zip.ZipException: invalid entry CRC

Hello,
We're trying to install Primavera EPPM 8.3 and test it on our server but having troubles in deploying p6 application.
We have successfully installed the following prerequisites:
- WebLogic 11g
- JDK1.7.0_45 and setting JAVA_HOME path=C:\Program Files\Java\jdk1.7.0_45 in enviornmental variables
- Downloading wsdl4j-1.6.2.jar for web services
- Running the setup.exe file (located in PEPPM path\Disk1\install\setup.exe
- Successfully finished the setup and running the Configuration Wizard.
- Successfully created the database using Microsoft SQL Server 2008 database
- Successfully creating a new WebLogic domain.
- We didn't install BI publisher nor any content repository ( would this affect running the primavera application? )
- In the configuration wizard summary we can see everything is marked successfully except the "Start Web Logic" step it showed an error and a message indicating that we should start it manually ( however, the domain is successfully created ).
Next, we run the startWebLogic.cmd file in our new created domain, then we run startNodeManager.cmd from WebLogic common/bin directory.
We can access the WebLogic administration panel successfully but no "p6" application in the deployments.
So, we click "install" then open the p6 directory and deploy it without errors.
But when when starting the P6 server and then try to start p6 application, it failed with the following error message:
java.util.zip.ZipException: invalid entry CRC (expected 0xa0a67af0 but got 0xef73b950)
Can you please help us to solve this? Please note that every other deployment is running correctly except p6.

The cause of errors was a dying Westen Digital drive, specially the 48G partition reserved only for $ORACLE_BASE (/dev/sdb3 mounted on /opt/oracle).
A simple quick scan of unmounted partition (badblocks -v /dev/sdb3) reported more than thousand new bad blocks (compared to the last scan six months ago). Although I strongly believe, specially in the case of WDC drives, that the best utility to "repair" bad blocks is the one that opens a window and prints the message: "Go to the nearest shop and buy a new drive", I was very curious to prove my suspicion just on this drive. After zero-filling the partition with dd, then formatting it (mke2fs -cc) and mounting again, the 11g installation and database creation (on the same partition) successfully completed, performing fast and smoothly. To make sure it was not a casual event, I repeated the installation once again with full success. The database itself is working fine (by now). Well, the whole procedure took me more than four hours, but I'm pretty satisfied. I learned once again - stay away from Western Digital. As Oracle cannot rely on dying drive, my friend is going tomorrow to spend a 150+ euro on two 250G Seagate Barracudas and replace both WDC drives, even though the first drive seems to be healthy.
In general there is no difference between loading correct data from good disk into bad memory and loading the incorrect data from dying disk into good memory. In both cases the pattern is damaged. For everyone who runs into the problem similar to this, and the cause cannot be easily determined, the rule of thumb must be:
1. test memory and, if it shows any error, check sockets and test memory modules individually
2. check disks for bad blocks regardless of the result of memory testing
Therefore I consider your answer being generally correct.
Regards
NJ

Java.lang.SecurityException: Security: Invalid Subject: principals

I am getting the following exception intermittently:
java.lang.SecurityException: Security: Invalid Subject: principals=[XXX, Administrators]
What i am doing is, i have two weblogic servers both running Weblogic 10.0 and running on different domains, a war is deployed on one server (server A) which sends a message to queue on another server (Server B), now everything works but if i restart B then A throws the above Security Exception while looking up the queue on Server B?? Any ideas why, i haven't configured any security credentials.
If i restart A after restarting B then everything works again but restarting all the servers each time one gets restarted is cumbersome,so does someone knows answer to the question above?
Edited by: user4828945 on Feb 11, 2009 5:41 PM

If you dont require authentication, then enable the global trust between the domains.
When this feature is enabled, identity is passed between WebLogic Server domains over an RMI connection without requiring authentication in the second domain. When inter-domain trust is enabled, transactions can commit across domains. A trust relationship is established when the Domain Credential for one domain matches the Domain Credential for another domain.
By default, the Domain Credential is randomly generated and therefore, no two domains will have the same Domain Credential. If you want two WebLogic Server domains to interoperate, you need to replace the generated credential with a credential you select, and set the same credential in each of the domains.
Link :[http://e-docs.bea.com/wls/docs100/ConsoleHelp/taskhelp/security/EnableGlobalTrustBetweenDomains.html]

Clear Metadata Cache

I had an error that I have listed below and had to clear the metadata cache to fix it. I do not have the metadata cache activated under user option so I am not sure why I had to clear the cache to get things working. Can someone explain why I had to do this and where the metadata cache is located?
Error:
Index was out of range. Must be non-negative and less than the size of
the collection.
Parameter name: startindex

Hello Jeroen,
There is no other possibility. If you adjusted and reimported a RFC model, you need to restart the J2EE server.
Regards,
Christophe

URGENT HELP about java.util.zip.ZipException: invalid entry CRC

I have a program (JAVA of course) packet on JAR with fat-jar eclipse plugin. This program work well in all my computers except two. On these two computers I receive a java.util.zip.ZipException: invalid entry CRC.
Both computers have the last version of java, but one is Windows and the other is Linux.
Any help to find the source of this problem??
Thanks in advance.

Sorry, I give poor information about this problem.
This is the full error showed when I execute this command: java -jar app.jar
Unable to load resource: java.util.zip.ZipException: invalid entry CRC (expected 0x358054d7 but got 0x7dc370ba)
java.util.zip.ZipException: invalid entry CRC (expected 0x358054d7 but got 0x7dc370ba)
at java.util.zip.ZipInputStream.read(Unknown Source)
at java.util.jar.JarInputStream.read(Unknown Source)
at java.io.FilterInputStream.read(Unknown Source)
at com.simontuffs.onejar.JarClassLoader.copy(JarClassLoader.java:818)
at com.simontuffs.onejar.JarClassLoader.loadBytes(JarClassLoader.java:383)
at com.simontuffs.onejar.JarClassLoader.loadByteCode(JarClassLoader.java:371)
at com.simontuffs.onejar.JarClassLoader.loadByteCode(JarClassLoader.java:362)
at com.simontuffs.onejar.JarClassLoader.load(JarClassLoader.java:305)
at com.simontuffs.onejar.JarClassLoader.load(JarClassLoader.java:224)
at com.simontuffs.onejar.Boot.run(Boot.java:224)
at com.simontuffs.onejar.Boot.main(Boot.java:89)
Exception in thread "main" java.lang.ClassNotFoundException: com.intarex.wizard.IWizard
at com.simontuffs.onejar.JarClassLoader.findClass(JarClassLoader.java:497)
at java.lang.ClassLoader.loadClass(Unknown Source)
at java.lang.ClassLoader.loadClass(Unknown Source)
at com.simontuffs.onejar.Boot.run(Boot.java:240)
at com.simontuffs.onejar.Boot.main(Boot.java:89)
app.jar is a JAR file created with fat-jar eclipse plugin, to make easier to generate a JAR file with all dependencies.
I think that is not a code related problem, because this program is executed in several computers without errors.
I trasport this JAR to the client computer via HTTP.
I'm trying to find clues to find the origin of this problem.
Thanks.

Signed jar: java.util.zip.ZipException: invalid entry size

Hi, I have written an applet using Java 1.4. The applet contains multiple .class and .gif files. I jared the files to produce a single .jar file with the manifest. Then I signed that jar file with jarsigner that produced a self-signed jar with the certificate. When I run the applet in IE6 on my computer (Windows XP) it works fine - I got the certificate dialog box popping up and the applet loads normally. My computer does not have a Web server installed so everything happens locally.
As soon as I move the html and signed jar files to the server and try to call it from my PC, the Java console displays the following message:
java.util.zip.ZipException: invalid entry size (expected 1342177310 but got 7728 bytes)
     at java.util.zip.ZipInputStream.readEnd(ZipInputStream.java:363)
     at java.util.zip.ZipInputStream.read(ZipInputStream.java:142)
     at sun.plugin.cache.CachedJarLoader.decompress(CachedJarLoader.java:397)
     at sun.plugin.cache.CachedJarLoader.access$500
CachedJarLoader.java:53)
     at sun.plugin.cache.CachedJarLoader$5.run(CachedJarLoader.java:335)
     at java.security.AccessController.doPrivileged(Native Method)
etc... it's huge
I have been through the forums, some people have similiar problem while trying to sign their jar, well, mine signs OK, it's just the fact that it doesn't run ... I've tried to kill the manifest and then re-sign the jar. I also tried to jar the files with -0 option - nothing worked. I do appreciate if someone could point me in the right direction.
Thanks

I have managed to work this one out: UNIX servers don't like jar files FTPed using ASCII. They should should be transferred in binary mode. I have installed an FTP client on my PC, switched to binary mode and re-transferred the jar file and it worked straight away.
Thanks.

WD Java RFC metadata cache invalidation without recycle JAVA

Similar Messages

Maybe you are looking for