Comms Express problem after installing patch - Oracle Communications Messaging Exchange Server

I had the mail and address book components of comms express working but was hitting the bug with the calendar not being able to show dates after 2006. So I installed patch 122793-12 which is supposed to fix this issue.
After installing the patch I got the following error on restarting the application server (which is the web container):
[#|2007-06-13T18:44:00.848+0100|SEVERE|sun-appserver-ee8.1_02|org.apache.catalin
a.session.ManagerBase|_ThreadID=10;|IOException while loading persisted sessions
: java.io.WriteAbortedException: writing aborted; java.io.NotSerializableExcepti
on: com.sun.uwc.common.util.UWCPreferences
java.io.WriteAbortedException: writing aborted; java.io.NotSerializableException
: com.sun.uwc.common.util.UWCPreferences
This occurs when the app svr tries to load the comms express web app.
I went through the uwcconfig.properties and uwcauth.properties carefully, comparing the old version and the version modified by the patch. The only change of note is that the patch added the following to uwcconfig.properties:
webmail.securedproxyauth=false
webmail.proxyadmin=ZZZ
webmail.proxyadminpass=ZZZ
I found that if I commented out these lines the application server started up correctly. However, on logging into comms express I found the Mail tab had dissapeared and clicking the Calendar tab produced "The selected calendar(s) was either deleted or does not exist or you do not have permissions to view it. Select another calendar(s).".
So any ideas as to (a) what these new webmail.proxy settings are all about and/or (b) any other ideas as to why this patch has messed up Comms Express?
The version of our Messenging Server is v6.2 Rev. 3.04 Build 2005.196.0138 and we are using Identity SSO.
Thanks. 

you need to uncomment those lines, and put the correct entries in. . . . 

Hi,
So any ideas as to (a) what these new webmail.proxy settings are all about and/or (b) any other ideas as to why this patch has messed up Comms Express?Please note that by installing patch 122793-12, you have actually upgraded to a new release of the product which introduced a number of underlying structural changes (one of these being that comms-express now communicates differently for webmail access).
Prior to doing this upgrade there are a large number of pre-requisite and also post-install steps that need to be carried out as per the upgrade guide:
http://docs.sun.com/app/docs/doc/819-7561
Other 6.3 (comm-suite-5) upgrade patches are available at:
http://www.sun.com/bigadmin/hubs/comms/downloads/updates.jsp
Regards,
Shane. 

Shane, thanks for alerting me to the fact that I had inadvertantly upgraded from 6.2 to 6.3. I decided to backout this patch and to install one of the obsoleted patches that fixes the calendar server date problem without upgrading to 6.3.
So I backed out the patch using backout-newconfig and restarted the application server.
The problem is that now when I try and login to communications express I get the following errors in the log:
SEVERE: Could not load the file: <uwcbasedir>/WEB-INF/domain/<Our.Domain.Name>/uwcdomainconfig.properties (No such file or directory)
Jun 27, 2007 3:22:02 PM com.sun.uwc.common.util.UWCDomainCache loadUWCDomainProp
erties
WARNING: Unable to load domain configuration file for domain<Our.Domain.Name>
uwcdomainconfig.properties has also just existed in the root of <uwcbasedir>/WEB-INF/domain and this directory has never had our domain name as a subdirectory. So why would installing/uninstalling this patch suddently cause Comm Express to start looking for things in different directories?
I have tried manually creating this directory and putting a copy of uwcdomainconfig.properties there but that caused "Invalid Session ID" errors.
I have checked uwcauth.properties and uwcconfig.properties and they are back to being exactly as they were before the patch was installed.
Thanks. 

"backout_newconfig" doesn't uninstall the patch, just the config changes. You also need to use the pkgrm command to remove the patch. 

Well, I used patchrm to remove the patch but I still get the same problem. 

are there error messages? Has anything changed since before you backed the patch out? "it still doesn't work" isn't really a very precise description of the problem. 

After uninstalling the patch it seems that Comms Express was expecting a domain specific configuration. I'm not sure why, but anyway, I created a directory for the domain under <uwcdeploydir>/WEB-INF/domain and copied the
uwcdomainconfig.properties and locale directories into it.
Then when I restart the application server there are no errors in either the app server log or uwc log. But when I browse to the Comms Express login page I get the following errors in the log:
FINE: IdentitySSOAuthFilter: Resolved domain? <My.Domain.Name>
Jun 28, 2007 11:35:27 AM com.sun.uwc.common.auth.IdentitySSOAuthFilter doFilter
FINE: 1. session is not present.. get token manager..
Jun 28, 2007 11:35:27 AM com.sun.uwc.common.auth.IdentitySSOAuthFilter doFilter
WARNING: SSOException..Invalid session ID.
Jun 28, 2007 11:35:27 AM com.sun.uwc.common.auth.LDAPAuthFilter doFilter
FINE: Check If Identity is enabled
Jun 28, 2007 11:35:27 AM com.sun.uwc.common.auth.LDAPAuthFilter doFilter
FINE: Identity is enabled....return
Jun 28, 2007 11:35:27 AM com.sun.uwc.common.util.UWCDomainCache loadUWCTheme
WARNING: Theme for localeenand<My.Domain.Name> is not configured
Jun 28, 2007 11:35:27 AM com.sun.uwc.common.util.UWCDomainCache loadUWCTheme
Nothing has changed other than the fact that I installed and uninstalled the patch. I have checked the uwcauth.properties and uwcconfig.properties and they are identical to the original versions (prior to installing the patch).
I have tried creating a locale and domain specific theme but that didn't help and I think the errors about the theme just result from the original "Invalid Session ID" error.
Thanks.

Related

Comms express / UWC repeatedly asking for login?

Building a test setup with a new front end machine to our msging/cal servers. Jes 2005Q1 release.
All going well except UWC deployed on the frontend machine won't let us log in - if ou get the passwd correct it just presents the login page again (it is authing as if you give an incorrect passwd it complains).
We have cal express and msg express deployed on this machine and AM based SSO is working fine across those.
Before I spend hours tomorrow searching for the config that I have missed has anyone got any clues where to start looking?
uwc.log gives:
Jun 8, 2005 5:27:49 PM com.sun.uwc.common.util.UWCDomainManager loadDomainFromIS
SEVERE: Exception..LDAP authentication failed.
Jun 8, 2005 5:27:49 PM com.sun.uwc.common.util.UWCDomainManager loadDomainFromIS
SEVERE: Exception..LDAP authentication failed.
Jun 8, 2005 5:27:49 PM com.sun.uwc.common.util.UWCDomainManager loadDomainFromIS
SEVERE: Exception..LDAP authentication failed.
Jun 8, 2005 5:27:49 PM com.sun.uwc.common.auth.IdentitySSOAuthFilter doFilter
FINE: 1. session is not present.. get token manager..
Jun 8, 2005 5:27:49 PM com.sun.uwc.common.auth.IdentitySSOAuthFilter doFilter
WARNING: SSOException..LDAP authentication failed.
Jun 8, 2005 5:27:49 PM com.sun.uwc.common.auth.IdentitySSOAuthFilter removeIdentityCookie
Cheers,
Darren 
Well, um, there are so many ways that UWC can fail.
The error messages quoted seem to imply you're not getting LDAP authentication.
"we have cal express and msg express" on "this" machine. You also talk about "frontend machine"
I'm now a little confused about your architecture.
UWC requires to either be installed on the same box as your back-end, or have MEM (Messaging Express Multiplexor), on the same frontend box. Does your MEM on this box work? 
Well, um, there are so many ways that UWC can fail.Tell me about it ;-)
The error messages quoted seem to imply you're not
getting LDAP authentication.thats what I thought but as I say, if I give the incorrect passwd it complains. If I give the correct passwd it gives no error so ldap authentication is working to some level.
"we have cal express and msg express" on "this"
machine. You also talk about "frontend machine"
I'm now a little confused about your architecture.
UWC requires to either be installed on the same box
as your back-end, or have MEM (Messaging Express
Multiplexor), on the same frontend box. Does your
MEM on this box work?Sorry, I was a bit vague!
Front end box has MEM and UWC on it. logging into MEM or the cal express running on this front end machine works fine. SSO is also working between these two clients. Only UWC is showing this problem.
I guess I was hoping that someone would recognise the symptoms - I'm not at the machines at the moment but I suspect I have missed some setting while attempting to follow the adventure game that is the JES3 docs :)
Cheers,
Darren 
Ah, better. . .
I would go back, and double-check all the configutil settings for your MEM and back-end. UWC requires some special settings, there, as I vaguely remember. sigh. Documentation could be better. . . 
We had a similar problem Some users could authenticate through uwc but my account could not. If I typed in my correct password I went back to the login screen with no message. If I typed in the wrong password I got the error string. I traced it to the fact that my LDAP entry had a multivalued cn. I removed one of the cn values and could authenticate and get to my mailbox 
Ok, not sure what the problem was but there was a nagging feeling that something was "not quite right" about the identity install on this machine.
Removed and reinstalled AM SDK (which wouldn't reinstall untill we manually removed a load of java demo packages....grrrrr) and all is now well.
Not sure what had happened - all very odd.
Thanks for the help anyway!
Darren 
I think that this is pretty common.
You have to be amazingly consistant about your ldap entries, all along the way.
You also have to install not just the Identity SDK, but the container, too. 
yeah. "SDK and config web container" or similar.
Do you know a way around the problem that when I uninstalled the ID server I couldn't then reinstall do to problems with existing packages? It wouldn't reinstall until I had manually removed a load of things.
We had to remove:
SUNWjato
SUNWjatodmo
SUNWjatodoc
SUNWjaf
SUNWjmail
SUNWjhrt
SUNWjhrt.2
before the installer would reinstall ID SDK which was annoying to say the least...
Cheers,
Darren 
Well, the bugs that are in the system (to be fixes) include the broken-ness of the uninstaller, and the broken-ness of the installer.
The uninstaller should remove all the correct packages. It doesn't.
The installer should be able to install over left-over packages. It isn't.
Getting the installer/uninstaller fixed is, alas, lower priority than getting the products themselves fixed.

Problem accessing Calendar in UWC

Dear All
I just installed the JES2005/Q4 on single server and everything is almost fine exclude calendar via UWC, calendar express working without any problem and also I can login with calmaster user name, but when I login to UWC, I can't access the Calendar and uwc leave the below logs inside uwc.log I check and doubled checked everything I tried to reinstall the server but still same problem, I also reconfigured the uwc few times but still same error, however I need to say my UWC is working under application server 8.
Dec 1, 2006 6:36:32 PM com.sun.uwc.common.UWCUserHelper createCalStore
SEVERE: calsession not created calstore connect has failed
Dec 1, 2006 6:36:32 PM com.sun.uwc.calclient.CalModuleServlet onInitializeHandle
r
SEVERE: Error[onBeforeRequest:getCalStore] [Error:getCalStore] - Could not creat
e store
any help is appreciated.
Best regards
Mo 
Couple of things you could check
1. The value of caldb.calmaster = "root#india.sun.com" (Maybe UNKNOWN) -Highyl likely
2. If in Virtual domain ,then check in uwcauth.properties and ics.conf if there is any misconfiguration.
3.Verify your password in ics.conf ,uwcauth.properties &,uwcconfig.properties.
thanks
Ramya

Problem with installed Opensso

Hello,
I’ve installed Sun OpenSSO Enterprise on Solaris 10 with Sun Directory Server, had it configured and had been using for about a month. Then, yesterday, I’ve rebooted the server. Now, when I go to the URL where OpenSSO was installed, I get the Configuration Option page. Does anyone have any idea how I can resolve this without having to reinstall everything from scratch and losing all the configurations?
Thanks,
Jen 
I've resolved this issue. It was because Opensso domain was started before directory server.
Edited by: angelmoon on Apr 21, 2009 2:43 PM 
In 7.1 this was usually due to the fact of the bootstrap file being gone.
Recreating the bootstrap file and pointing to to your config would fix it.

Register directory V7.0 fails on dsee7

This is a brand new installation of DSEE v7.0 64bit on RedHat 5.x 64bit.
Everything seems to be working okay. I created several directory instances manually, and they all run fine. However, when I try to register an instance within dscc7 I get an error in step 1.2: Review Server Certificate.
A window pops up saying: An error occurred trying to acept the certificate. The error is com.sun.directory.dcc.ads.ADSContextException.
I would appreciate any help in this matter.
Thanks!
Enrique.
Edited by: Enrique on Jul 13, 2010 5:20 PM 
Most likely the hostname you used to create the instance does not match the system's actual hostname. 
Thanks akillenb!
I also found an article stating that it could have been an issue with space under ..../dscc7var/tmp. I checked the filesystem, and it was really low, only a few kb left. I added space to the file system, ran the registration, and Voila! Everything worked okay I do not remember having to change the hostname, it showed up automatically...
Enrique.

Can't log into Virtual Desktop Connector Admin GUI?

See below for specific version info:
Virtual Desktop Connector 1.0 Install/Admin guide says that after running vda-config you log into the admin GUI originally as the Solaris root user and root password but it apparently does not work? There were other documentation errors in the install guide and just wondering if the default login information is incorrect as well?
I get no errors in the logs. The only error is on the admin gui login page which simply indicates that Authentication Failed and that the user name/ and or password is wrong, try again.
Of course I know my root password and can log onto the system with that password and no I'm don't have cap locks on or anything like that. I think it's the documentation and that there is a default admin user/password that should be used other than root.
Here is what I have installed:
- VMware ESX and VCenter 3.5 U3 but that should not be affecting me from simply logging into the VDC admin GUI.
- Sun Ray Server 4.0 w/the latest Solaris X86 patch 127554-05
- Sun Secure Global Desktop 4.41
- Sun Ray Connector for Windows 2.0
and finally the suspect piece:
- Sun Virtual Desktop Connector 1.0 w/latest Solaris X86 patch 127560-03
All installs went smoothly with no errors. Everything but VDC has been tested and is working as expected. Now I'd just like to log into the admin GUI and configure VDC.
Both vda-service status and vda-webadmin status show as running with an associated pid.
When running vda-config I chose the defaults in all cases:
- JRE = /usr/java
- Apache-tomcat = /opt/apache-tomcat and it is version 5.5.20 (at least 5.5 is supposedly required)
- Using local files for datastore versus using SGD or SRSS datastore. Documentation says it is the default and is fine for demo purposes, though you might want to use SGD or SRSS datastore when rolling it out in production.
- Webadmin is set for remote access and the apache-tomcat user id is the default vdawww account which is created as I see it in /etc/passwd.
?? Everything seems to be fine. I just can't log into the VDC Admin GUI with the default root user and password which the documentation states is to be used.
Thank You, 
It's working for me so I'm trying to remember if I had any problems with it. I think it asks you if you want to enable it during the install but I don't see that in the install log so maybe that was the other version. First thing I can thing of is maybe your trying to log into the wrong interface... Are you going to port 1800 or 1801( if using ssl ) on your box? I know for my demo I put SGD and the VDC on the same box so maybe you're trying to log into SGD?
Either way, I don't think it's a documentation thing because it's definitely working for in my setup. Not sure if it makes a difference but I picked Evaluation version when I installed mine.
I 'think' that this is the log for the admin gui... Not sure if that will help. /var/opt/SUNWvda/log/webadmin0.log 
I am also running in eval mode. I have VDS, SGD and SRSS all on the same machine but I'm definitely connecting to the VDC interface at 1801 in secure mode.
The log in screen actually says Sun Virtual Desktop Connector.
The last things in my webadmin0.log state:
Feb 26, 2009 4:24:51 PM org.apache.coyote.http11.Http11BaseProtocol start
INFO: Starting Coyote HTTP/1.1 on http-1800
Feb 26, 2009 4:24:51 PM org.apache.coyote.http11.Http11BaseProtocol start
INFO: Starting Coyote HTTP/1.1 on http-1801
Feb 26, 2009 4:24:51 PM org.apache.catalina.startup.Catalina start
INFO: Server startup in 7240 ms
All normal from what I can see.
Did you choose (3) to use the local file system for the data store or did you select to use SGD's or SRSS's data store?
Do you happen to know where the file is which lists the VDC administrator and/or password?
Thanks 
I can't find it in the install logs but I'm almost positive that I used the SGD data store. I don't know where it stores the password. It seems to be running an LDAP directory for things by looking at the install log, so I would guess either there or it's using PAM? 
I am having this problem every time after rebooting the VDI-Server.
but I found out that after restarting the cacaoadm
#cacaoadm stop --force
#cacaoadm start
I can log in with root and password.

Categories

Resources