Friday 31 August 2012

Installing WLS (web logic server) 10.0.3.5 properly for BSSV

WLS install

download jrockit-jdk1.6.0_33-R28.2.4-4.1.0-linux-x64.bin from oracle

put it on your linux machine

install it (graphical or text based installation)

I put it in /u01/app/oracle/jrockit

use it to run the wls installer (need a graphical session)

image

/u01/app/oracle/jrockit/bin/java –jar wls1035_generic.jar

image

yay – no OUI (Oracle universal installer)

Also, delete is rm –Rf on the install dir and u can install again, no hacking the

 

image

The screen about staying informed about updates is tough and takes quite a few clicks (choosing the wrong items)

Choose advanced, and get rid of coherence

image

Next screen is important, use jrockit!!!

image

next, next next…  should only take 30 seconds to install

Choose YES to run quick start

image

Click the highlighted option

image

create new

next on select domain source

image

See above, I like to use machine name in the name of the domain, because when it comes up in EM12C, it just has the domain name…

image

next

image

Must be production mode for BSSV!

image

Choose both top tick boxes

image

Dont’ check SSL

Next on managed servers

next on clusters

image

Choose unix tab and add a name (of your machine)

image

NExt

image

Move the admin server over under your machine

Then create

takes 8 seconds

Done!

manage from http://vsydwls01:7001/console

start with startNodeManager.sh (find it under base install)

start admin server with startWebLogic.sh (find it – under domain)

You need need to create your contain and also install the BSSV server from SM now… then when it will not deploy, follow the rest!

Now, hack the config.xml for your current domain (find . -name config.xml –print)

for me  /u01/app/oracle/Middleware/user_projects/domains/osl_vsydwls01_domain/config/config.xml

back it up

search for <server-start>

directly below that line, create your own <class-path> entry

<class-path>/u01/app/oracle/Middleware/wlserver_10.3/server/adr/ojdbc6.jar</class-path>

Save and exit

restart AdminServer

deploy BSSV – job done!

Installing WLS (web logic server) 10.0.3.5 properly for BSSV

WLS install

download jrockit-jdk1.6.0_33-R28.2.4-4.1.0-linux-x64.bin from oracle

put it on your linux machine

install it (graphical or text based installation)

I put it in /u01/app/oracle/jrockit

use it to run the wls installer (need a graphical session)

image

/u01/app/oracle/jrockit/bin/java –jar wls1035_generic.jar

image

yay – no OUI (Oracle universal installer)

Also, delete is rm –Rf on the install dir and u can install again, no hacking the

 

image

The screen about staying informed about updates is tough and takes quite a few clicks (choosing the wrong items)

Choose advanced, and get rid of coherence

image

Next screen is important, use jrockit!!!

image

next, next next…  should only take 30 seconds to install

Choose YES to run quick start

image

Click the highlighted option

image

create new

next on select domain source

image

See above, I like to use machine name in the name of the domain, because when it comes up in EM12C, it just has the domain name…

image

next

image

Must be production mode for BSSV!

image

Choose both top tick boxes

image

Dont’ check SSL

Next on managed servers

next on clusters

image

Choose unix tab and add a name (of your machine)

image

NExt

image

Move the admin server over under your machine

Then create

takes 8 seconds

Done!

manage from http://vsydwls01:7001/console

start with startNodeManager.sh (find it under base install)

start admin server with startWebLogic.sh (find it – under domain)

Now, hack the config.xml for your current domain (find . -name config.xml –print)

for me  /u01/app/oracle/Middleware/user_projects/domains/osl_vsydwls01_domain/config/config.xml

back it up

search for <server-start>

directly below that line, create your own <class-path> entry

<class-path>/u01/app/oracle/Middleware/wlserver_10.3/server/adr/ojdbc6.jar</class-path>

Save and exit

restart AdminServer

deploy BSSV – job done!

Thursday 23 August 2012

debugging BSSV deploy

step 1: get your logging right

edit d:\jdedwards\e900\system\classes\jdelog.properties

Ensure that you use a complete path, relative ones are hard to find c:\\jde.log

 

step 2: work out which config files are being read

Full nuts logging:

#[LOG2]
FILE=c:\\jasdebug.log
LEVEL=DEBUG
FORMAT=TOOLS_THREAD
MAXFILESIZE=10MB
MAXBACKUPINDEX=20
COMPONENT=ALL
APPEND=TRUE

23 Aug 2012 20:41:13,327 [Line -1] [Thread-3] [APP   ]  - [5008][BASE]            Loading config from file D:\JDEdwards\E900\PY900\ini\sbf\jdbj.ini per the default_path system property

This tells you the files that are being read.

Modify them, then goto mysupport.oracle.com and look for

E1:BSSV: Troubleshooting Package Deployment Issues with Business Services (BSSV) [ID 1400146.1]

installing EM12C JD Edwards management pack

pwd

Download the code from edelivery

https://edelivery.oracle.com/EPD/Download/get_form?egroup_aru_number=14255326

V28759-01.zip

This is actually an opar file - 12.1.0.1.0_oracle.apps.jded_2000_0.opar

oracle documentation on the process is here http://docs.oracle.com/cd/E24705_01/doc.91/e26776/toc.htm

Note that I’ve already installed EM12C and IMPORTANTLY configured the software library!

What you expect to see is JD Edwards appear when you goto your setup->extensibility->plugins menu in EM12C

image

as above

Note that you’ll need to sftp / scp the upgrade (zip) file to the EM12C console machine

you’ll also need a command line shell to the EM12C console machine to complete the steps below

[oracle@myrem12c e112cAgent]$ unzip V*
Archive:  V28759-01 (1).zip
extracting: 12.1.0.1.0_oracle.apps.jded_2000_0.opar

To import the JD Edwards Application Pack OPAR, use this sequence of line commands:

  1. Change to the oms/bin directory of your EMCC installation. For example:

    cd /u01/app/emgc12/oms/bin/


  2. Enter this command to set up the import of the JD Edwards EnterpriseOne Application Pack into EMCC: (this is your em12c login URL)

    ./emcli setup -url=https://myrem12c:7799/em -username=sysman -password=Oracle123 -trustall


  3. Verify the console display indicates the command was successful as shown in this example:

    Oracle Enterprise Manager Cloud Control 12c Release 12.1.0.0.0.
    Copyright (c) 1996, 2011 Oracle Corporation and/or its affiliates. All rights reserved.

    Emcli setup successful


  4. Run this command to perform the import of the JD Edwards EnterpriseOne Application Pack OPAR into EMCC:


    [oracle@myrem12c bin]$ ./emcli import_update -file=/u01/e112cAgent/12.1.0.1.0_oracle.apps.jded_2000_0.opar -omslocal
    Processing update: Plug-in - Oracle Jdedwards EnterpriseOne Plugin consists of monitoring and management for Oracle Jdedwards EnterpriseOne system.
    Operation completed successfully. Update has been uploaded to Enterprise Manager. Please use the Self Update Home to manage this update.



  5. At this point the import of the JD Edwards EnterpriseOne Application Pack OPAR into EMCC is complete.


To go any further, you need to make sure that you have some agent installs on a machine that is part of your JD Edwards infrastructure.  You are going to deploy the opar to one of your agents so that you can start to monitor things.


Log into the EM12C console and goto setup->plugins


Choose the newly installed “Oracle Jdedwards Enterpriseone” plugin under the “Applications” folder.  Deploy on, Management server.  


Note that this is going to restart OMS – so proceed only if you can handle this!


image


image


 


You now need to deploy it to an agent machine, hopefully one that has some component of JDE on it.


Once there, you need to add it as a system, goto


emcli get_plugin_deployment_status -plugin_id=oracle.apps.jded


Next will be deploy to an agent.  You need to select a machine in your JDE infrastructure on which to install the management agent update.  This is done from the pulgin’s page.  Choose the JDEdwards EnterpriseOne Application and “deploy on” Management Agent.


image


choose the machine you want to use.


image


 


goto targets –> systems


image


choose JD Edwards from the pull down and Add


image


fill out the details like below:


image


image


Make sure that you’ve started the agent on the e1 machine that you are using as the Agent host machine!  Note also that the lookup for Enterprise Manager Agent Host Name does not work, so just type in the value.


image


You’ll then have your JDE system listed and ready to manage!

more about deploying EM12C–manually

I love things when they work, why…  well – they work.  I like the fact that my estimates are accurate and also that I don’t have to go working out things that I’ll never need to know again.  Yet, uch of my blogging is inspired by things that do not work!

As in the current case.

I’ve been trying to deploy an new em12c agent for some time now – what a pain!  I had a em12c agent in this machine previously, but it was not BP1.  I thought that I’d remove the old and install the new – as you do.  I was having all sorts of bother (well, it did all start when I did an rm –rf from the agent install dir – then processes to hack the inventory.xml in the contentsXML dir in the oracle inventory).  Yeah, I know what you are saying, this is all my fault.

Anyways, I kept getting errors with the installer about not being able to connect back to the genwallet

Validating oms host & port with url: http://myrem12c.mits.local:4901/empbs/genwallet
Validating oms host & port with url: https://myrem12c.mits.local:4901/empbs/genwallet
Return status:1
WARN: Forcing the configuration of the agent when the OMS is not available will configure the agent to use HTTP (non-secure) communication.  In order to properly establish secure HTTPS communication between the agent and the OMS, you will need to secure the agent after the OMS is available.

All my firewalls were off, so it was annoying.  I decided to cut and paste the install command from the EM12C window to my agent host and see if I could solve it manually.

 

/u01/app/em12c/ADATMP_2012-08-22_20-14-24-PM/agentDeploy.sh -ignorePrereqs ORACLE_HOSTNAME=e1ent2.mits.local AGENT_BASE_DIR=/u01/app/em12c OMS_HOST=myrem12c.mits.local EM_UPLOAD_PORT=4901 AGENT_INSTANCE_HOME=/u01/app/em12c/agent_inst b_doDiscovery=false b_startAgent=false b_forceInstCheck=true AGENT_PORT=3872

The command above was still valid, as the cleanup from the aborted install had not occurred

Remember you can add –forceConfigure, and the Return status:1, is just a warning…

/u01/app/em12c/ADATMP_2012-08-22_20-14-24-PM/agentDeploy.sh -ignorePrereqs ORACLE_HOSTNAME=e1ent2.mits.local AGENT_BASE_DIR=/u01/app/em12c OMS_HOST=myrem12c.mits.local EM_UPLOAD_PORT=4901 AGENT_INSTANCE_HOME=/u01/app/em12c/agent_inst b_doDiscovery=false b_startAgent=false b_forceInstCheck=true AGENT_PORT=3872

So again, BAM I let it have it…  Got a bunch of SAX XML parsing errors…  Wait…  I know what that means…  Troubleshooting 101…  I have an XML document problem…  Where do I remember something about that?  OOh yeah, hacking the inventory.xml file…

Once I tidied up the </HOME> tag in the inventory.xml, I was away!  Here is the output in case it helps you:

Validating the OMS_HOST & EM_UPLOAD_PORT
Executing command : /u01/app/em12c/core/12.1.0.1.0/jdk/bin/java -classpath /u01/app/em12c/core/12.1.0.1.0/jlib/agentInstaller.jar:/u01/app/em12c/core/12.1.0.1.0/oui/jlib/OraInstaller.jar oracle.sysman.agent.installer.AgentInstaller /u01/app/em12c/core/12.1.0.1.0 /u01/app/em12c/ADATMP_2012-08-22_20-14-24-PM /u01/app/em12c -prereq

Validating oms host & port with url: http://myrem12c.mits.local:4901/empbs/genwallet
Validating oms host & port with url: https://myrem12c.mits.local:4901/empbs/genwallet
Return status:1
WARN: Forcing the configuration of the agent when the OMS is not available will configure the agent to use HTTP (non-secure) communication.  In order to properly establish secure HTTPS communication between the agent and the OMS, you will need to secure the agent after the OMS is available.
Unzipping the agentcoreimage.zip to /u01/app/em12c ....
12.1.0.1.0_PluginsOneoffs_226.zip
Executing command : /u01/app/em12c/ADATMP_2012-08-22_20-14-24-PM/unzip -o /u01/app/em12c/ADATMP_2012-08-22_20-14-24-PM/12.1.0.1.0_PluginsOneoffs_226.zip -d /u01/app/em12c
Executing command : /u01/app/em12c/core/12.1.0.1.0/jdk/bin/java -classpath /u01/app/em12c/core/12.1.0.1.0/oui/jlib/OraInstaller.jar:/u01/app/em12c/core/12.1.0.1.0/oui/jlib/xmlparserv2.jar:/u01/app/em12c/core/12.1.0.1.0/oui/jlib/srvm.jar:/u01/app/em12c/core/12.1.0.1.0/oui/jlib/emCfg.jar:/u01/app/em12c/core/12.1.0.1.0/jlib/agentInstaller.jar:/u01/app/em12c/core/12.1.0.1.0/oui/jlib/share.jar oracle.sysman.agent.installer.AgentInstaller /u01/app/em12c/core/12.1.0.1.0 /u01/app/em12c/ADATMP_2012-08-22_20-14-24-PM /u01/app/em12c AGENT_BASE_DIR=/u01/app/em12c -ignorePrereqs ORACLE_HOSTNAME=e1ent2.mits.local AGENT_BASE_DIR=/u01/app/em12c OMS_HOST=myrem12c.mits.local EM_UPLOAD_PORT=4901 AGENT_INSTANCE_HOME=/u01/app/em12c/agent_inst b_doDiscovery=false b_startAgent=false b_forceInstCheck=true AGENT_PORT=3872 -forceConfigure


Cloning the agent home...
Executing command: /u01/app/em12c/core/12.1.0.1.0/oui/bin/runInstaller -ignoreSysPrereqs -clone -forceClone -silent -waitForCompletion -nowait ORACLE_HOME=/u01/app/em12c/core/12.1.0.1.0   AGENT_BASE_DIR=/u01/app/em12c ORACLE_HOSTNAME=e1ent2.mits.local AGENT_BASE_DIR=/u01/app/em12c OMS_HOST=myrem12c.mits.local EM_UPLOAD_PORT=4901 AGENT_INSTANCE_HOME=/u01/app/em12c/agent_inst b_doDiscovery=false b_startAgent=false b_forceInstCheck=true AGENT_PORT=3872 -noconfig  ORACLE_HOME_NAME=agent12c1 -force b_forceConfigure=true
Clone Action Logs Location:/u01/app/oraInventory/logs/cloneActions<timestamp>.log
Cloning of agent home completed successfully

Attaching sbin home...
Executing command: /u01/app/em12c/core/12.1.0.1.0/oui/bin/runInstaller -ignoreSysPrereqs -attachHome -waitForCompletion -nowait ORACLE_HOME=/u01/app/em12c/sbin ORACLE_HOME_NAME=sbin12c1 -force
Attach Home Logs Location:/u01/app/em12c/core/12.1.0.1.0/cfgtoollogs/agentDeploy/AttachHome<timestamp>.log
Attach home for sbin home completed successfully.

Updating home dependencies...
Executing command: /u01/app/em12c/core/12.1.0.1.0/oui/bin/runInstaller -ignoreSysPrereqs -updateHomeDeps -waitForCompletion HOME_DEPENDENCY_LIST="{/u01/app/em12c/sbin:/u01/app/em12c/core/12.1.0.1.0}" -invPtrLoc /u01/app/em12c/core/12.1.0.1.0/oraInst.loc -force
Update Home Dependencies Location:/u01/app/em12c/core/12.1.0.1.0/cfgtoollogs/agentDeploy/UpdateHomeDeps<timestamp>.log
Update home dependency completed successfully.

Performing the agent configuration...
Executing command: /u01/app/em12c/core/12.1.0.1.0/oui/bin/runConfig.sh ORACLE_HOME=/u01/app/em12c/core/12.1.0.1.0 RESPONSE_FILE=/u01/app/em12c/core/12.1.0.1.0/agent.rsp ACTION=configure MODE=perform COMPONENT_XML={oracle.sysman.top.agent.11_1_0_1_0.xml} RERUN=true
Configuration Log Location:/u01/app/em12c/core/12.1.0.1.0/cfgtoollogs/cfgfw/CfmLogger<timestamp>.log

Agent Configuration completed successfully

The following configuration scripts need to be executed as the "root" user.
#!/bin/sh
#Root script to run
/u01/app/em12c/core/12.1.0.1.0/root.sh
To execute the configuration scripts:
1. Open a terminal window
2. Log in as "root"
3. Run the scripts
Agent Deployment Successful.
Agent deployment log location:
/u01/app/em12c/core/12.1.0.1.0/cfgtoollogs/agentDeploy/agentDeploy_<timestamp>.log
Agent deployment completed successfully.

That was very satisfying.  I now know that I could do bulk agent deploys with the same scripts and know-how…  Nice!

Remember to run the root script (/u01/app/em12c/core/12.1.0.1.0/root.sh)

Tuesday 14 August 2012

create a basic linux printer for use with JD Edwards–OEL linux

In JD Edwards, it’s easy.  P98616, add new printer, done.

At the command line it’s a little more challenging.  What I mean is, using linux commands to get the printing working is a little more of a challenge.  If you have a gnome-session, then you can probably start the printer configuration application from System->Administration->printers.

When you add the printer, choose a decent printer name, description and location.  Use App Sockey / HPDirect (if you want to print direct to the IP address of the printer).  I decided to so this, rather than to use the windows print servers.  You could use the samba option if this is the case.  Note that my title is basic printer.

Then it’s simply choosing the machine name (make sure that you can ping it – well, at least arp it of ping is not enabled).

lpadmin is how the hardcore create the printer.

you need to find a driver that is going to work, the RAW one will not.

lpinfo –m will list all of the printers and the drivers.  You are going to need one of these PPD files to give to the command line to create the printer.

I advise to grep for something generic like

lpinfo –m |grep "HP LaserJet 5 Foomatic" | grep recommended

foomatic:HP-LaserJet_5-hpijs.ppd

Time to create your printer!

lpadmin -p myPrinter –D"Money Printer" –E –L "Lunch Room HQ" –v"socket://sydxerox53.osl.local:9100" -m"foomatic:HP-LaserJet_5-hpijs.ppd"

This will create a linux printer called myPrinter

show all printers

lpstat –a

show all outstanding jobs

lpstat –l

Monday 13 August 2012

starting nodeManager and startWebLogic automatically

There is a lot of stuff on the new, but not a lot of good stuff on this topic.  My server is in production mode, so you cannot just pass is usernames and passwords to the start web logic script.  Perhaps you can use a here file, but that might be for next time.  I tried a basic here file and did not get too much success.  I read the script startWebLogic.sh and thought that the only option was to set the WLS_PW and WLS_USER values for the shell that starts the command.  The problem with this method is that when you do “ps –ef |grep weblogic”, you get a bit of a surprise!

I’ve done a couple of things to get this working.  Firstly I’ve edited the nodemanager.properties file to start things automatically if there is a crash.  Not too sure if this is going to help if the adminServer is already down?  I thought it was going to be the fix that I was waiting for…  Not too sure that it was.

edit the file below

./wlserver_10.3/common/nodemanager/nodemanager.properties

change the lines to be true

CrashRecoveryEnabled=true
StartScriptEnabled=true

I did a restart of the node manager and the admin server did not come up – Doh.

So, I then created the following script called weblogic and whacked it into the /etc/init.d spot: 

remember to chmod 700, as I have some secret squirrel passwords in it. 

It’s a bit fancy, and I attribute the fanciness to BP who works with me.  You know who you are, thanks!

 

#!/bin/sh
#
# /etc/rc.d/init.d/weblogic
#
# A start/stop script for weblogic on Oracle Unbreakable Linux x86_64
#
# chkconfig: 2 99 1
# description: Starts and stops weblogic
#

# Settings and initialisation.
SERVICE="weblogic"
DESCRIPTION="weblogic and notemanager"
SETCOLOR_TITLE="echo -en \\033[0;36m"
RETVAL=0

# Service defaults
WLSUSER="oracle"
RUNWLS="/bin/su ${WLSUSER} -lc"
WLSPROJ="/u01/app/oracle/Middleware/user_projects/domains/my_domain"
WLSSVR="/u01/app/oracle/Middleware/wlserver_10.3/server"
WLS_USER=weblogic
WLS_PW=**Your weblogic user password”


# Source function library.
if [ -f /etc/init.d/functions ] ; then
  . /etc/init.d/functions
elif [ -f /etc/rc.d/init.d/functions ] ; then
  . /etc/rc.d/init.d/functions
else
  exit 1
fi

# Avoid using root's TMPDIR
unset TMPDIR

start() {
        echo -n $"Starting ${DESCRIPTION} services: "
        echo
        RETVAL=0

        echo ; $SETCOLOR_TITLE ; echo "weblogic:" ; $SETCOLOR_NORMAL
        ${RUNWLS} "export WLS_USER=$WLS_USER;export WLS_PW=$WLS_PW; ${WLSPROJ}/bin/startWebLogic.sh &2>1 &"
        RETVAL_WLS=$?
#        sleep 60

        echo ; $SETCOLOR_TITLE ; echo "weblogic node manager:" ; $SETCOLOR_NORMAL
        ${RUNWLS} "${WLSSVR}/bin/startNodeManager.sh &2>1 &"
        RETVAL_WLSNM=$?
#        sleep 60

        RETVAL=`expr $RETVAL_WLS + $RETVAL_WLSNM`
        [ $RETVAL -eq 0 ] && touch /var/lock/subsys/${SERVICE} || RETVAL=1

        return $RETVAL
}

stop() {
        echo -n $"Shutting down ${DESCRIPTION} services: "
        echo
        RETVAL=0

        echo ; $SETCOLOR_TITLE ; echo "weblogic:" ; $SETCOLOR_NORMAL
        ${RUNWLS} "${WLSPROJ}/bin/stopWebLogic.sh $WLS_USER $WLS_PW &"
        RETVAL_WLS=$?

        kill -9 `ps -ef |grep -i "jdk Weblogic.nodemanager" |grep -v grep | awk '{print $2}'`

        RETVAL=`expr $RETVAL_WLS`

        [ $RETVAL -eq 0 ] && rm -f /var/lock/subsys/${SERVICE}
        return $RETVAL
}


status() {

                echo ; $SETCOLOR_TITLE ; echo "${DESCRIPTION} weblogic Status:" ; $SETCOLOR_NORMAL
                ps -ef|grep startWebLogic.sh|grep -vE 'grep|ps -ef|-sh|awk|sort|uniq|ksh|su'|awk '{print $2, $9, $10, $11, $12, $13}'
                echo ; $SETCOLOR_TITLE ; echo "${DESCRIPTION} weblogic node manager Status:" ; $SETCOLOR_NORMAL
                ps -ef|grep startNodeManager.sh|grep -vE 'grep|ps -ef|-sh|awk|sort|uniq|ksh|su'|awk '{print $2,$9}'
}

case "$1" in
  start)
        start
        ;;
  status)
        status
        ;;
  stop)
        stop
        ;;
  *)
        echo $"Usage: $0 {start|status}"
        exit 2
esac

exit $?

 

Despite being a little verbose, the above script is working!

[root@vsydwls01 init.d]# chkconfig weblogic --add
[root@vsydwls01 init.d]# chkconfig  --level 5 weblogic on
[root@vsydwls01 init.d]# chkconfig  --level 3 weblogic on
[root@vsydwls01 init.d]# chkconfig weblogic --list
weblogic        0:off   1:off   2:on    3:on    4:off   5:on    6:off

 

What do those level numbers mean: (thanks Wikipedia).  3 & 5 are very common choices.

Code Information
0 Halt
1 Single-User mode
2 Multi-user mode console logins only (without networking)
3 Multi-User mode, console logins only
4 Not used/User-definable
5 Multi-User mode, with display manager as well as console logins (X11)
6 Reboot

Thursday 9 August 2012

Bulk insert into project (all custom tables)

This is a quick work around for a scenario that I have found a number of times.  The predicament I was currently in… (I was on a bus, travelling at 60km/h – no – just kidding)   I needed to copy custom tables from AS/400 to oracle.  But..  I did not have a fat client which had the specs for the custom tables.  F98403 needs the specs, so I was in trouble.  There was about 70 custom tables, so I did not want to manually pop them all into my project for an advanced get (to get the specs to my deployment server).  Oh yeah, my enterprise server was linux too, so there was no ODBC to run the conversions on the ent server… So, I ran the following:

INSERT INTO SY900.F98222 (POOMWPRJID, POOMWOBJID, POOMWOT, POSRCRLS, POOMWUSER, POOMWMKEY, POENHV, POPATHCD, PODATS, POOMWCHS, POOMWOST, POOMWOVS, POOMWOBSDT, POOMWCRTDT, POOMWAC, POOMWTC, POOMWTCV, POOMWPOS1, POOMWPOS2, POOMWPOS3, POOMWPOS4, POOMWPOS5, POOMWPOD1, POOMWPOD2, POOMWPOD3, POOMWPOD4, POOMWPOD5, POOMWPON1, POOMWPON2, POOMWPON3, POOMWPON4, POOMWPON5, POPID, POMKEY, POUSER, POUPMJ, POUPMT, POTIMEZONES, PODSAVNAME)
SELECT 'GenerateTables', siobnm    , 'TBLE' ,'E900','JDE','VSYDJDE02','PY900','PY900',' ','0','01',' ',0,0,' ',' ',' ','0',' ',' ',' ',' ',0,0,0,0,0,0.000000,0.000000,0.000000,0.000000,0.000000,'P98220','DSBSLSMOIRS','MOIRS',104303,143243.000000,' ',' '
FROM ol900.f9860 where siobnm like 'F55%';
commit ;

My project name was GenerateTables.  I was release E900.  Did an advanced get, ran the R98403.  Everyone was happy!

Wednesday 8 August 2012

More MSDE / SEE problems for 9.1 tools release install

I finally go the database being created, but then started getting “login prompts” for “PKgBuild MSDE …” – wow is this going to end?  This is in relation to the SA password and the fact that JD Edwards needs an encrypted password beyond 8.98.4.2.  I tried hard to use the system\bin32\reconfigureMSDE.exe – but could not get it working.  It seems that my SA account was unlocked.

image

I have no access to SQLServer tools, so the following did it!

C:\Users\xxx>osql -S VSYDJDE02\JDESSELOCAL -E
1> alter login sa with password = 'S0mething99!' UNLOCK
2> go
1> quit

Then plug that password into the reconfigureMSDE.exe and finally the build might start working.

Package build frustrations–SSE

I was getting the following errors in the clientPkgBuild.log file:

Wed Aug 08 09:10:57  -  PkgMisc.cpp:1926 PKGMSC0014 ERROR: Failed to create MSDE spec repository.
Wed Aug 08 09:10:57  -  Repository created/attached.
Wed Aug 08 09:10:57  -  pkgSpec.cpp:2018 PKGSPC0013 ERROR: Failed to create spec repository on the build machine.
Wed Aug 08 09:10:57  -  Detaching the local repository on the build machine. This will take some time.
Wed Aug 08 09:10:57  -  Detaching local repository SPEC__PY900OSL.

If you follow [ID 867954.1] it’ll get you to uninstall and reinstall SSE – but probably to no avail (like in my situation).  My fix was actually to run the commands below:  Although, I did not get the error message that was alluded to in the logs that oracle point to.

It’s a little frustrating that you need to enable debug to get logs like the ones below:  Surely these are log worthy!

4676/3576 WRK:Starting jdeCallObject            Wed Aug 08 09:10:57.869003    MSDEDBAPI.cpp993
    MSDE000013 - SQLCreateDatabase failed for 'SPEC__PY900OSL'

4676/3576 WRK:Starting jdeCallObject            Wed Aug 08 09:10:57.869005    MSDEDBAPI.cpp1124
    [Microsoft][Microsoft SQL Server Driver]CREATE DATABASE failed. Primary file must be at least 3 MB to accommodate a copy of the model database. SQLSTATE: 42000 HRESULT:0X80040E14  Native Error:1803.

4676/3576 WRK:Starting jdeCallObject            Wed Aug 08 09:10:57.869009    spkgutil.c527
    [PKGBLD] - Wed Aug 08 09:10:57  -  PkgMisc.cpp:1926 PKGMSC0014 ERROR: Failed to create MSDE spec repository.

 

c:\>sqlcmd -E -S (local)\JDESSELocal
1> dbcc shrinkdatabase (model)
2> go

Tuesday 7 August 2012

give me table names and row counts (for a schema) in a table oracle syntax

I do this all of the time.  When I’m converting data, doing upgrades or just for fun!  I need to list all of the tables in an owner that have data.  I know there are 10000 ways of doing this, but this is what generally works when I use oracle or SQLServer or AS/400.  With a slight tweak of syntax!

create table ol900.dataaudit (tablename varchar(20) not null, rowcount integer) ;

run the results of the following:  This is going to get row counts for ALL tables in CRP

select 'insert into ol900.dataaudit (select ' || '''' || table_name || '''' || ', count(1) from crpdta.' || table_name || ');' from all_tables where owner = 'CRPDTA';

Look at your results.

select * from ol900.dataaudit order by rowcount desc ;

tool release 9.1 on apps release 9.0 deployment server–which install program do I use? Which oracle home to choose.

What a nightmare, you can only get this right with lots of reading and some pretty difficult searching.    The succinct version is below. 

backups first (system, systemcomp and oneworld client install)

In the link http://docs.oracle.com/cd/E24902_01/nav/installation.htm, under the deployment server reference guide e18836.pdf, you can find the steps to install the Tools release for the deployment server.


The correct setup.exe to use is under folder ../ToolsRelease/Install  (D:\software\9104DepServer\Disk 1\ToolsRelease\install\setup.exe in my example)

image

specify home details – JD Edwards enterprise one screen – choose your existing home.  Note that if OUI was not used for the 8.98 install (which I can tell you it was not).  The home that you want to upgrade will not exist.  So create a new one (like above) and point to the E900 dir on the deployment server (you got backups?)

 

Hit yes, don’t be scared!

image

next – go!

image

Then hopefully you get:

image

Friday 3 August 2012

my simple guide to RIA–Related Information Architecture

This is hard to find information on!

use this document first Related Information Architecture

p952332 – work with related information

Add

image

Fields make sense if you know what RIA is about.

Tables used:

  • Generic URL Metadata Table (F952334)
  • Generic URL Access control Table (F952335)
  • Related Information Cross-Reference Header Table (F952330)
  • Related Information Cross-Reference Detail Table (F952331)
  • General RIA info (F952332)

image

Note that mine are missing a *ALL in the version field…  SQL time

update sy900.f952332 set bhbhVER = '*ALL' where BHBHAPPL = 'P03B31' ;

ok – so once this was updated (the version field)

image

We have RIA working perfectly, immediately.

deploying your first JAS server to new WLS and there are NO machines?

image

What a pain, don’t stress though.  The fix is super fast.

You’ll probably need to start your admin server, this will not start “out of the box” on a new install of wls

go searching for “startWebLogic.sh” in your middleware home dir.  You’ll find it a couple of times under the domain that you created.  Hopefully you’ve created one!

run the copy of startWebLogic.sh in bin32

Log into the weblogic console, probably with a URL something like:

http://webserver:7001/console

login with the weblogic user and password that you entered on install (if you are using an e1 server template I think its ovsadminE1)

You’ll get this totally awesome interface

image

Note that if you are ex-OAS – you might fall off your chair this interface is that cool!

goto environment, machines

image

press lock and edit

image

New

image

In our context, a machine is the logical representation of a computer that hosts one or more weblogic instances, so it makes sense to call your machine by it’s hostname!

image

Finish

image

Just remember your port number

image

Done, activate the changes

image

Bosch:  your server will now appear!  You’ll need to start the create new container wizard again in SM.

image