WebIAppletStartup not found and Apache/Tomcat Split deployment

[Former Member]

I'm experiencing strange issue with the Split deployment of Web part of SAP BO 4.0. I used the following pdf meterial 'Improving the User Experience in SAP BI Platform 4.0 with Apache and WDeploy' written by James Rapp.

My Environment consists of:
Default Tomcat
Apache 2.2- the latest release, downloaded from vendor's web
SAP BO 4 with Fp3

what I did:
1. modified Tomcat's xml file by adding <Connector port="8009" protocol="AJP/1.3" redirectPort="8443" URIEncoding="UTF-8" enableLookups="false" connectionTimeout="20000"/>
2. Downloaded approprirate mod_jk file from web, uploaded to Apache modules folder
3. Created workers file & modifid the default httpd config file as it's written in PDF guide
4. Executed wdeploy.bat tomcat6 –Das_mode=split –Dws_type=apache deployall

Result: everything is fine - all web interfaces are up and running, except WebIntelligence from BI protal, here I'm getting the following error
when I start WebIntelligence.

load: class com.sap.webi.ui.jrp.WebIAppletStartup not found.

java.lang.ClassNotFoundException: com.sap.webi.ui.jrp.WebIAppletStartup
at sun.plugin2.applet.Applet2ClassLoader.findClass(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadClass0(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadClass(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadClass(Unknown Source)
at java.lang.ClassLoader.loadClass(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadCode(Unknown Source)
at sun.plugin2.applet.Plugin2Manager.createApplet(Unknown Source)
at sun.plugin2.applet.Plugin2Manager$AppletExecutionRunnable.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)

The default installation with Tomcat is fine, WebIntelligence interface is also running as expected, but when I modified it by adding Apache then this
error message occurs. Maybe someone can help me out, what can be cause of the error message.

I did some search in SAP Notes are, but did not find any reasonable explanation. For me it seems that the split deployment if ok, just Java is failing for some unknown reason....

Thanks in advance

[jmsrpp · Accepted solution]

Hi Ainars,

This appears to be a regression in FP03 as I have reproduced the problem internally using FP03 Patch 4.  I have requested your colleague raise a support message to address this and provide me with the message number.  I will ensure a reproduction environment is available to the assigned engineer so we can get this fixed in a patch.

Please let me know if you have any further questions and thanks very much for raising this to our attention.  We'll do our best to get this addressed as soon as possible, and I have confirmed with your colleague that there is a temporary workaround while we get this fixed.  The workaround is simply to copy the timestamp based folder that exists within $htdocs/BOE/portal so that the 0000000 directory exists as well.

[Former Member · Accepted solution]

Hi James,

The issue seems to exist in BI 4.0 SP04 as well. I did the full install on Linux using WebLogic 10.3.6 and Apache 2.2 for the front end. The BI 4.0 apps are built in a split mode. The Web Intelligence Application failed to open with the errors similar to one above. I could overcome it only using a workaround you suggested - copy the timestamp based folder that exists within $htdocs/BOE/portal so that the 0000000 directory exists as well.

Regards,

Inna

[jmsrpp · Accepted solution]

Hi Inna,

It does indeed exist in SP04 since that code originated from the FP03 release.  I was able to have this fixed in Patch 4.3 under ADAPT01639394.  You may use the workaround or consider patching to 4.3 or higher.

I've checked the forward fit rules (https://service.sap.com/bosap-maintenance-schedule) and it looks like this is also fixed in SP05. 

In summary, you may use the workaround or apply Patch 4.3+ or SP05.

Regards,

Jim

[Former Member · Accepted solution]

Hi Experts,

Sorry to revive an old thread but I am having the same behavior where I am unable to modify any webi report when I go through the Apache---Tomcat route. I get the same error as above, and unable to find a fix. Should i be redeploying the files?

Any help is appreciated.

[jmsrpp · Accepted solution]

Hi,

This thread mentions both a workaround and a fix that should be available in Patch 4.3 and SP05.  Have you followed those processes?

[Former Member · Accepted solution]

Hi James,

Thanks for your quick reply. I am on SP4 Patch2 so was trying to go the workaround route however trying to understand that do you want me to rename the timestamp folder to 0000000 from 1207301222? or am i missing anything

[jmsrpp · Accepted solution]

Hi,

In order to use the workaround it is necessary to make a copy of the timestamp folder in $htdocs/BOE/portal and rename it to 0000000.  This requires some additional disk space but will ensure that the request from Web Intelligence is served properly without the code fix.

[Former Member · Accepted solution]

Sure will do make a copy. Also should we be only renaming the BOE/PORTAL/Time stamp folder or all the other folders too which have the timestamp as not zeroes?

[Former Member · Accepted solution]

Hi James,

The rename didnt really fix the error. I still get the Webi Applet not found error. The logs look like: -

/BOE/portal/0000000/AnalyticalReporting/webiApplet/language/webi_swing_client_cdp_en.jar?version-id=14.0.4.738

Content-Length: 6,578

Content-Encoding: pack200-gzip

java.util.zip.ZipException: Not in GZIP format

at java.util.zip.GZIPInputStream.readHeader(Unknown Source)

at java.util.zip.GZIPInputStream.<init>(Unknown Source)

at java.util.zip.GZIPInputStream.<init>(Unknown Source)

at com.sun.deploy.net.HttpDownloadHelper.download(Unknown Source)

at com.sun.deploy.cache.Cache$4.run(Unknown Source)

at java.security.AccessController.doPrivileged(Native Method)

at com.sun.deploy.cache.Cache.downloadResourceToCache(Unknown Source)

at com.sun.deploy.cache.DeployFileOutputStream.close(Unknown Source)

at java.io.FilterOutputStream.close(Unknown Source)

at sun.net.www.protocol.http.HttpURLConnection$HttpInputStream.close(Unknown Source)

at java.io.BufferedInputStream.close(Unknown Source)

at sun.plugin.PluginURLJarFileCallBack.downloadJAR(Unknown Source)

at sun.plugin.PluginURLJarFileCallBack.access$000(Unknown Source)

at sun.plugin.PluginURLJarFileCallBack$1.run(Unknown Source)

at java.security.AccessController.doPrivileged(Native Method)

at sun.plugin.PluginURLJarFileCallBack.retrieve(Unknown Source)

at sun.net.www.protocol.jar.URLJarFile.retrieve(Unknown Source)

at sun.net.www.protocol.jar.URLJarFile.getJarFile(Unknown Source)

at sun.net.www.protocol.jar.JarFileFactory.get(Unknown Source)

at sun.net.www.protocol.jar.JarURLConnection.connect(Unknown Source)

at sun.plugin.net.protocol.jar.CachedJarURLConnection.connect(Unknown Source)

[jmsrpp · Accepted solution]

This is a different error.  I expect it is covered in this document and comments:

[Former Member · Accepted solution]

Hello James,

I have followed your instuctions here "Improving the User Experience in SAP BI Platform - BI 4.1 and Apache 2.4 Supplement" and other documents that relate to this subject matter. I am having the same issue in this post with BOE XI 4.1 SP4 when trying to launch the WEBI applet. Error message is below, do you have any suggestions? I have tried you workaround without any success. Thank you.

load: class com.sap.webi.ui.jrp.WebIAppletStartup not found.
java.lang.ClassNotFoundException: com.sap.webi.ui.jrp.WebIAppletStartup
at sun.plugin2.applet.Applet2ClassLoader.findClass(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadClass0(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadClass(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadClass(Unknown Source)
at java.lang.ClassLoader.loadClass(Unknown Source)
at sun.plugin2.applet.Plugin2ClassLoader.loadCode(Unknown Source)
at sun.plugin2.applet.Plugin2Manager.createApplet(Unknown Source)
at sun.plugin2.applet.Plugin2Manager$AppletExecutionRunnable.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)
Exception: java.lang.ClassNotFoundException: com.sap.webi.ui.jrp.WebIAppletStartup

[jmsrpp · Accepted solution]

Hi Kevin,

This almost certainly means that there is a missing AJP rule or you're still experiencing the issue I mention above that is covered in my Apache 2.4 supplement.

Please check the contents of your error.log and mod_jk.log to see what the detailed exception is.

Thanks,

Jim

[Former Member · Accepted solution]

I have gone over your post and others and still can not find a solution. Below is my Mod_jk.log

[Mon Sep 22 18:31:59 2014][5068:14352] [info] init_jk::mod_jk.c (3383): mod_jk/1.2.40 initialized [Mon Sep 22 18:31:59 2014][5068:14352] [info] init_jk::mod_jk.c (3383): mod_jk/1.2.40 initialized [Mon Sep 22 18:32:00 2014][11376:5752] [info] init_jk::mod_jk.c (3383): mod_jk/1.2.40 initialized [Mon Sep 22 18:32:00 2014][11376:5752] [info] init_jk::mod_jk.c (3383): mod_jk/1.2.40 initialized [Mon Sep 22 18:36:56 2014][11376:5144] [info] ajp_send_request::jk_ajp_common.c (1652): (ajp13) all endpoints are disconnected, detected by connect check (1), cping (0), send (0) [Mon Sep 22 18:37:49 2014][11376:5144] [info] ajp_send_request::jk_ajp_common.c (1652): (ajp13) all endpoints are disconnected, detected by connect check (2), cping (0), send (0) [Mon Sep 22 18:39:00 2014][11376:9588] [info] ajp_send_request::jk_ajp_common.c (1652): (ajp13) all endpoints are disconnected, detected by connect check (4), cping (0), send (0) [Mon Sep 22 18:39:22 2014][11376:14804] [info] ajp_send_request::jk_ajp_common.c (1652): (ajp13) all endpoints are disconnected, detected by connect check (3), cping (0), send (0)

and this is from the error.log

[Mon Sep 22 18:31:59.717583 2014] [mpm_winnt:notice] [pid 5068:tid 412] AH00455: Apache/2.4.10 (Win64) mod_jk/1.2.40 configured -- resuming normal operations [Mon Sep 22 18:31:59.717583 2014] [mpm_winnt:notice] [pid 5068:tid 412] AH00456: Apache Lounge VC11 Server built: Jul 17 2014 12:58:29 [Mon Sep 22 18:31:59.717583 2014] [core:notice] [pid 5068:tid 412] AH00094: Command line: 'C:\\Apache24\\bin\\httpd.exe -d C:/Apache24' [Mon Sep 22 18:31:59.717583 2014] [mpm_winnt:notice] [pid 5068:tid 412] AH00418: Parent: Created child process 11376 [Mon Sep 22 18:32:00.233195 2014] [mpm_winnt:notice] [pid 11376:tid 360] AH00354: Child: Starting 64 worker threads.

Any suggestons. Thank you for any and all help.

[jmsrpp · Accepted solution]

Theres bound to be an error in one of error.log or mod_jk.log.  If not mod_jk (which does look innocuous) then check error.log. 

[Former Member · Accepted solution]

and this is from the error.log

[Mon Sep 22 18:31:59.717583 2014] [mpm_winnt:notice] [pid 5068:tid 412] AH00455: Apache/2.4.10 (Win64) mod_jk/1.2.40 configured -- resuming normal operations [Mon Sep 22 18:31:59.717583 2014] [mpm_winnt:notice] [pid 5068:tid 412] AH00456: Apache Lounge VC11 Server built: Jul 17 2014 12:58:29 [Mon Sep 22 18:31:59.717583 2014] [core:notice] [pid 5068:tid 412] AH00094: Command line: 'C:\\Apache24\\bin\\httpd.exe -d C:/Apache24' [Mon Sep 22 18:31:59.717583 2014] [mpm_winnt:notice] [pid 5068:tid 412] AH00418: Parent: Created child process 11376 [Mon Sep 22 18:32:00.233195 2014] [mpm_winnt:notice] [pid 11376:tid 360] AH00354

[jmsrpp · Accepted solution]

I expect it would show up elsewhere if this is the case, but what about a 404 in access.log when you reproduce the problem?  If so, what's the offending URI?

[Former Member · Accepted solution]

No 404's but attached is what is in the access log when reproduce the error.

Something else I find strange. If I do not do the split and just add the following line "JkMount /* ajp13" to the http.conf along with what else is needed for mod_jk.so I get the same error message.

access.txt.zip

[Former Member · Accepted solution]

I got it to work after I removed the following two lines and there configuration from the http_bi41.conf. The two lines are LoadModule cache_module modules/mod_cache.so and LoadModule cache_disk_module modules/mod_cache_disk.so.

[jmsrpp · Accepted solution]

How interesting ... this means the issue was with an incorrectly cached response body from the original error.  Even though you fixed the underlying problem by following the workaround, the cached error body was still being returned to our request for the Webi Applet.

You can test this easily by re-adding the mod_cache related lines, deleting the full contents of the CacheRoot directory, and restarting Apache.  This would also explain why nothing showed up in the error log!

[Former Member · Accepted solution]

That did the trick. It is now working with mod_cache enabled.

[BillW · Accepted solution]

HI Jim,

I know about the time stamp folder. What I don't know is what file is this time stamp folder referenced in? The reason I ask is because I need to replace the new folder time stamp to the file. So it will point to the new folder.

Thanks
,

BW

[BillW · Accepted solution]

Ok I figured it out. It shows up in the URL and the access log. I thought there might be a place other then these two that it would be kept.

Thanks,

BW

[jmsrpp · Accepted solution]

Hey Bill,

It is an internal reference for Tomcat. In Apache it is an explicitly named folder. This is a byproduct of the build process for wdeploy. Apologies for missing the question but this thread is getting a bit lengthy.

Cheers,

Jim

[former_member190781]

Hi James,

I have been facing the same issue with split deployment in BI40 SP07. I tried the workaround but unfortunately it didnt work. The issue only occurs for BI Launchpad Webi modify preferences set as Rich Internet Application.

Appreciate your help.

Regards,

Sohel

[jmsrpp]

Hi Sohel,

See the previous post ... check for details.

[former_member190781]

Thanks James, I will go through it

Regards,

Sohel

[denis_konovalov]

This guide was writen/tested with SAP BI4.0 SP2 release, it should work with all BI4.x releases , but since you are using FP3 , which is in ramp up right now - you should follow on this with your ramp up coach and SAP support, as since FP3 is not a released product it might have some bugs/change in functionality in this area and we'll need to fix it quickly if that is the case.

[Former Member]

Dobrij denj Denis

Well, yes I do know that FP3 is brand new product and does hove a lot of changes since the release of  Fp02.. maybe you can add your expertise - what can cause this error message in a normal environment? I did multiple searches in Notes area, but did not manage to find anything useful.

To my understanding it states that one application value can not be downloaded for some reason, also I noticed that Monitoring is not available in Split deployment which also is quite strange.

Thanks in advance

[denis_konovalov]

try to reach to James, I'm not sure if he monitors this forums, but I'm sure he'll respond if you send him a message thru SCN functionality.

You mentioned that default tomcat is fine, maybe split deployment and default tomcat deployment are not worki8ng together, maybe you need pure split deployment on both tomcat and apache.... (tomcat BOE app is stripped)...

[jmsrpp]

It is a good point Denis.  I have not yet had an opportunity to test the split deploy on FP03 and it is possible there are issues there.

The customer did reach out to me directly and we have uncovered the following thus far:

[Tue May 15 07:41:03 2012]
[error] [client 127.0.0.1] File does not exist:
D:/Apache/htdocs/BOE/portal/0000000,

referer: http://localhost/BOE/portal/1205150648/AnalyticalReporting/WebiView.do?noDocument=true&cafWebSesInit...

[Tue May 15 07:41:03 2012]
[error] [client 127.0.0.1] File does not exist:
D:/Apache/htdocs/BOE/portal/0000000

[Tue May 15 07:41:03 2012]
[error] [client 127.0.0.1] File does not exist:
D:/Apache/htdocs/BOE/portal/0000000

[Tue May 15 07:41:03 2012]
[error] [client 127.0.0.1] File does not exist:
D:/Apache/htdocs/BOE/portal/0000000

[Tue May 15 07:41:03 2012]
[error] [client 127.0.0.1] File does not exist:
D:/Apache/htdocs/BOE/portal/0000000

WDeploy should create the static content directory within Apache using a timestamp format, such as:

1205121329

This folder name represents that WDeploy was executed at  2012, 05 (May), 12, 13:29

For some reason the mapping to Apache htdocs has become corrupt. 

We're planning to investigate further this week and I will update the thread with the results of our findings.