Windchill 10 Install - "specified user schema is not empty"

[egifford]

I'm setting up a test install of Windchill 10 on a pair of win 2008 R2 virtual servers (vmware). All was going well yesterday until the lastpart of running the PSI on the Windchill server (2 server setup - 1 for Oracle database, one for Windchill services &file vault). During the last steps the data loader failed and the PSI aborted (see attached screen shot). I tried some items suggested by users on PlanetPTCin an older post for similar errors involving windchill wt.auth.Authentication but wasn't successful. The original poster resolved by removing Windchill and re-running the PSI. In trying this,when running the PSI againI get to the point of specifying the Oracle user name for the Windchill installation and get the error "specified user schema is not empty" (see screen shot). My guess is the original run of the PSI established this user in the Oracle DB and the PSI can't create it again...but I'm just guessing. Is there a way to remedy this? I can always drop back and start fresh on the DB server as well, but I'd rather find a way to getfix what is going on with the PSI if possible. Thanks

Erik

[mmeadows]

Eric,

Looks like during the first installation, the installer made it past the point of Oracle database user creation and table population.

I assume you wiped out the first installation logs when you tried a second time. It would be useful to know what the original error was in C:\ptc\WINDCHILL\Windchill\installer\logs\WindchillLoader.log. For example, if the Oracle installation isn't complete (for Windchill), you will run into the same error again.

To recover from the second failure, drop and recreate your "pdmlink" user in Oracle. From a Windchill shell on the application server:

cd db/sql

sqlplus system/<password>@wind

drop user pdmlink cascade;

@create_user;

pdmlink

TEMP

USERS

quit

Now run the installer and tell it to populate an existing database user (pdmlink).

Regards,

Matt Meadows

Solutions Architect
VIRSO Inc
O: 618 937 8115
C: 314 749 8377
E: mmeadows@virsoinc.com

[JørnAHansen]

Erik,

Oops! Sorry for maybe misguiding you at http://communities.ptc.com/thread/33700

Probably what happened is that running the PSI on both the Oracle and the Windchill machine resulted in the PSI trying to load the schema 2 times.

I think following what Matt suggests above should bring you back on track again.

Jørn

[egifford]

I actually ended up deleting the database on the Oracle server using the Database Configuration Assistant anddeleting the Windchill install on the Windchill app / services server. After that re-running the PSI on the Oracle server to set up the database and user and then the PSI on the Windchill server to connect to the database got me past the user schema issue. However,the PSI still fails on the Windchill server at loading the base data. I've tried the "set ANT_OPTS=-Xmx512M:MaxPermSize=256M" fix suggested in the Windchill 10 RTF document, but it does not get me past the data load. Right now I'm working with PTC tech support to get this fixed and get 10 running. From other errors I'm getting I think I have a java issue.

Erik

[egifford]

As an example, this is what I get when I try to start windchill from a windchill shell (see screen shot).

Can anyone point me in a direction to check? I supplied PTC with install and tomcatlog files as requested yesterday but am not getting a response from them. I realize I'm loading pre-production software, but I know some of you out there have done this and been successful. I didn't have these issues when I installed 9.1 on virtual servers running Wiindows server 2003. At this point I'm not sure which variable is causing the problem - again, my test systems are VMWare virtual servers running Windows Server 2008 R2.

Thanks

Erik

[patrick.chin1]

I have a couple of suggestions based on what when this has happened to me:

• Block size is above the default 8K which is the old Oracle 9i/10G settings. Block sizes for 11.01.02 or 11.02.02 are set for 16K or you can set for even 32K. If that's the case, you just have to change your tiny block size values and perform the following:

• Modify your site.xconf

<property name="wt.generation.sql.tinyBlobSize" overridable="true"&lt;/span">

targetFile="codebase/user.properties"

value="64/>

<property name="wt.generation.sql.smallBlobSize" overridable="true"&lt;/p">

targetFile="codebase/user.properties"

value="64/>

• in a windchill shell
  • run xconfmanager -p
  • JavaGen.sh/bat registry false false true false false
• Change directory to%WT_HOME%/db/sqlordb/sql3 (if you have multi-byte char)
• sqlplussystem/<password>@<sid>
• drop the windchill oracle schema user
  • drop user <pdmlink> cascade;
  • @create_user;
  • <pdmlink>
  • TEMP
  • USERS
• connect <pdmlink>/<pdmlink>@<sid>
• spool create_pdmlink_schema.log
• @create_ddl_wt;
• @Make_DDL_install.sql;
• spool off;
• commit;
• exit

Re run your setup command with loading the previous install and select continue. It should load the database schema.

Another possibility is If your issue is that you have Oracle 11.02.02,Windchill 10may not uselatest oracle JDBC drivers and still be using the 11.1.0.7.0. It will result in this issue if the database schema failing which will be shown in your $WT_HOME/installer/logs/*WindchillDDL.log(s). I know that Windchill 9.1 M060 has the right oracle JDBC drivers to connect for loading. Windchill 9.1 M050 does not while loading the database. You can start with 11.02.01 to get it working. You can then export and import to a 11.02.02 database and that works while using Windchill 9.1 M050 or your Windchill 10.

That's the only possible issues I know of that may result in your issue. But, you haven't included details of your logs in the failure.

[egifford]

I had someone email asking if I found a fix to this, sorry for the delay in posting back. I did get Windchill 10 working. I applied the Windchill 10 Preproduction QAP patch and my install startedbehaving properly. I was able to run a manual data load after applying the patch (both auto as part of using the PSI and a manual load would fail before) so I'm now up and running with 10 on test servers. What is unclear is whether the QAP patch actually included a specific fix for the issue(s) I was encountering or if running the patch install corrected something else sort of secondarily. The information with the patch lists a slew of files that are replaced / updated by the patch but no descrption of what errors it fixes.

Anyway, if you run into the same problem I was having try applying the QAP patch if you haven't. I was expecting 10 to at least install properly and run before the patch and just "run better" after the patch. PTC tech support said nothing abouttheQAPwhen I called them with the install issues I was having, so I was trying to fix what was fundamentallynot workingbefore applying a patch. If anyone else has the same problem and it is fixed after the QAP please post back. Or if someone looks at what's in the patch and can deduce an"aha moment" where clearlyone of the replaced files would have been the fix, let me know. In these cases I like to know why something is fixed, not just accept that it suddenly goes from broken to running.

Thanks!

Erik

In Reply to Erik Gifford:

As an example, this is what I get when I try to start windchill from a windchill shell (see screen shot).

Can anyone point me in a direction to check? I supplied PTC with install and tomcatlog files as requested yesterday but am not getting a response from them. I realize I'm loading pre-production software, but I know some of you out there have done this and been successful. I didn't have these issues when I installed 9.1 on virtual servers running Wiindows server 2003. At this point I'm not sure which variable is causing the problem - again, my test systems are VMWare virtual servers running Windows Server 2008 R2.

Thanks

Erik