Setting up the Script Based Node Manager

I noticed that someone has already asked this question from the Java side but I was wondering about the process for the node manager used for Solaris. I have been through the process in the documentation but I have yet to complete the dots to get a working nodemanager. The node manager starts up but then the admin console is unable to start down servers. I can shut a managed server down but I cannot bring it back up unless I physically (well virtually) log into each WLS and start them from the command line as root (yes, I have some servers running in privileged port ranges). I got it so when it starts, the servers switch users from root to my weblogic user. I just cant figure out the rest of the configuration.
Any feedback here?
Here is what I do:
1. Installed Weblogic and configured my domain.
2. I setup my SSL Trusts between boxes.
3. Set the weblogic nodemanger password and port.
4. Configured the nodemanager for each managed server in the console (I am totally unsure if I did this right or not).
5. Ran NMEnroll on all of my wls machines including the AdminServer.
6. Started the NodeManger on all machines.
7. Tested and was unable to control any of the managed servers from the console from a shutdown state.

I think I figured out whats going on. I think maybe I confused the two types of node managers. It appears to me that the way the script based one works is that you give it a command and then it uses a SSH shell connection to issue the command. Here is my case, I am on Solaris 10 and I have servers running on ports in the privileged range. I ran the test in the manual and it will not startup. The command works, but I am being stopped by the OS. How do you modify the script node manager to run as root over SSH when you need to sudo? I would normally consider setting up sudo to run without a password only for the weblogic server startup commands. However, I know in a production system the security folks would never permit this. Right now I can stop servers but I cannot start them remotely without manually SSHing in and using sudo to start the managed servers.

Similar Messages

Migrate to Java Based Node Manager

Hi all ,
We are planning to migrate from script based node manager to Java based node manager . Can any one suggest the procedure or detailed steps in doing the same . If you have any documents then its really helpful .
thanks,
Kiran

Hi,
I couldn't find direct way of saying whether node manager is running or not but here is the work around. Using WLST when you say "startNodeManager", if the node manager is running this command throw an output saying "node manager is already running". Let me know if this solution works for you, if it works then I can give you java program for this if required or you can use "WLST JAVA Mode".
Thanks.
Vijay Bheemineni.

Setting up the months based on prompts.

Hi All,
My requirement is that I am selecting the calender year(2009) and calender month(8) in the prompts I should be able to retrieve the previous month and previous to previous month in the per_name_month fashion that is (2009 /07) which I have to use in the pivot table.
What are the ways to do it?

I think I figured out whats going on. I think maybe I confused the two types of node managers. It appears to me that the way the script based one works is that you give it a command and then it uses a SSH shell connection to issue the command. Here is my case, I am on Solaris 10 and I have servers running on ports in the privileged range. I ran the test in the manual and it will not startup. The command works, but I am being stopped by the OS. How do you modify the script node manager to run as root over SSH when you need to sudo? I would normally consider setting up sudo to run without a password only for the weblogic server startup commands. However, I know in a production system the security folks would never permit this. Right now I can stop servers but I cannot start them remotely without manually SSHing in and using sudo to start the managed servers.

What is the use of Node Manager?

OBIEE Oracle Business Intelligence 11.1.1.5
I have started up the WebLogic server which delivers the WL Server Admin Console as well as the Fusion Middleware Control.
I have ignored NodeManager. It hasnt been started.
What is the use of NodeManager?
Oracle tells us:
Note: In order to start the Managed Server using the WebLogic Server Administration Console, the Node Manager must be running.
http://download.oracle.com/docs/cd/E21043_01/bi.1111/e10541/components.htm
But my Manged Server is running ok - I can check http://<host>:7001/console and its ok.
So why do I need to bother with NodeManager?
I only have WLS and OHS installed - no BI instance.
DA

Answers inline:
As I have installed weblogic server 10.3.5 on windows. while am starting it using cmd am getting two ips as
1**.*.... for weblogic server one is (channel default 4... .. and one is default 1) kindly explain.
- It doesn't generate any IP. If you have installed successfully, weblogic will run on your system and it's ip will remain same. You can access it using your systems'IP:portnumber
Also tell as I have to use cmd for starting servers how can i use unix commands for the same??
- You have installed on windows so unix command won't work for you. You can either start by going into services window in windows or start using command by going to path weblogichome\domains\domainname\bin. Execute startWebLogic.cmd
Also explain what is the use of node manager and weblogic scripting tool which are also installed in the system.
- Node manager is used for communication bentween two or more servers in a domain.
As am looking to install OIM 11g so kindly tell what are all the things I need to know about weblogic so that i can learn OIM easily.
- To install OIM, install weblogic, install oracle database. Use RCU utilities to create database and then install SOA, OIM and configure them. You can get all kinds of guides and steps for OIM 11g r2 here:
http://docs.oracle.com/cd/E27559_01/index.htm
regards,
GP

Why MicroSoft 2000 SR needed while running the script using test manager???

hi all,
whenever I trying to run a script using test manager,one POP-UP comes which starts installing MICROSOFT 2000 SR.
and in few seconds it complaining for some missing component,
can any body tell me why it is happening.
what is the need of Microsoft 2000 SR while running the script in test manager.
tnx
USOni

USoni
I have seen that problem happening before, even that i was never able to figure out why, something that you can try is changing the service log in credentials for both:
Oracle Application Testing Suite Agent Service
Oracle Application Testing Suite Application Service
and use a proper user with admin rights.
let me know if that helped
Regards
Alex

Registering with the WebEx Data Center and the Cisco WebEx Node Management System

Dear guys, ...
Please help,
i want to implement to webex node ASR1000, i have read in "Configuring the Cisco Webex Node for ASR 100.pdf", there is prerequisites to implement it, that is "Registering with the WebEx Data Center and the Cisco WebEx Node Management System"
Can someone tell how to "Registering with the WebEx Data Center and the Cisco WebEx Node Management System"
Are there any step by step documentation to "Registering with the WebEx Data Center and the Cisco WebEx Node Management System"?
Thank you
BR

You should have received a PAK Key with your order. Go to Cisco licensing and enter the PAK Key as this will start the process. Once the PAK Key is validated a screen will be displayed to enter your request for ASR 100 integration. It normally takes a few days to a couple of weeks to get the information back from WebEx needed to configure your ASR.
If you did not get a PAK Key contact your WebEx rep to get the process started to integrate your ASR to your WebEx site.
Hope this helps
John

How to conrol the size of Node Manager error logs

How to control the size of Node manger error.logs in Weblogic 8.1?

Log rotation is disabled by default, but can be enabled by setting LogLimit and LogCount in nodemanager.properties
For more information check [http://download.oracle.com/docs/cd/E11035_01/wls100/nodemgr/java_nodemgr.html]
I hope this helps

What is the function of Node Manager in WLS 8.1 SP4

Hi,
I was trying to understand the real function of Node Manager. During my install I had install Node Manager to run as Windows Service (w2k server). I was wondering as to how i can setup Node Manager to monitor my manage servers (from what i know it can do) and can it restart any manage servers that stop and/or are not started.
Thanks
J

This should help http://e-docs.bea.com/wls/docs81/adminguide/confignodemgr.html
cheers !
Dips

What is Node Manager ?

Hi
I have been reading documents on Oracle Fusion Middleware 11g. At lot of places 'Node Manager' is mentioned across these documents.
Please can someone explain here what is Node Manager and what it is its role in OFMW 11g ?
Cheers
Mayur

Hi Mayur,
For more details please refer to: http://weblogic-wonders.com/weblogic/2010/06/18/weblogic-initials/
 What is NodeManager?
A Node Manager process is not associated with a specific WebLogic domain but with a machine. Node Manager process to control server instances in any WebLogic Server domain until the server instances reside on the same machine as the Node Manager process.
Node Manager is a WebLogic Server utility that enables you to start, shut down, and restart Administration Server and Managed Server instances from a remote location. Although Node Manager is optional, it is recommended if your WebLogic Server environment hosts applications with high availability requirements.
A Node Manager process is not associated with a specific WebLogic domain but with a machine. You can use the same Node Manager process to control server instances in any WebLogic Server domain, as long as the server instances reside on the same machine as the Node Manager process.
Why We Need Nodemanagers?
1). It can Start, Shut Down, and Restart an Administration Server which has assigned a Machine.
2). It can Start, Shut Down, and Restart an Managed Server which has assigned a Machine.
3). It can Restart Administration and Managed Servers. (Crash Recovery of Servers is an Interesting feature)
4). It can be used to Monitor Servers and View Log Data
Types of NodeManagers?
*1). Java Based Nodemanager:* This can be used in Windiows as well as UNIX based Operating systems. Java Based Node Manager is not supported on Open VMS, OS/390, AS400, UnixWare, or Tru64 UNIX.
*2). Script Based Nodemanager:* This nodemanager can be used in UNIX based operating systems. The advantage of the script-based Node Manager is that it can remotely manage servers over a network that has been configured to use SSH.
For More Information on NodeManager please refer to:
http://download.oracle.com/docs/cd/E13222_01/wls/docs90/server_start/nodemgr.html
And
http://blogs.oracle.com/jamesbayer/2010/01/weblogic_nodemanager_quick_sta.html
Thanks
Jay SenSharma

Node Manager Not Reachable

Hi Everyone,
I'm very new to WebLogic, so I apologize if this is a dumb question.
I'm trying to install WebLogic as server for OBIEE. My environment is pretty simple. I have two Oracle Red Hat Linux 32-bit physical machines, let's call them WLS_SERVER and OBIEE_SERVER. I installed OBIEE on the OBIEE_SERVER and tested it using the out of the box OC4J container and that worked fine.
Then I installed WebLogic 11g on the WLS_SERVER machine and I ran the config.sh script. I used the configuration GUI to register the two physical machines and an additional server object called BIServer. Out of the box, web logic comes with AdminServer server object. I heard it's a bad practice to deploy applications to the AdminServer, hence why I created the BIServer object. I associated BIServer with the OBIEE_SERVER machine.
I went to start the BIServer and I got the following error:
"For server BIServer, the Node Manager associated with machine OBIEE_SERVER is not reachable."
How do I setup a Node Manager on the OBIEE_SERVER? Do I also need to physical put the WebLogic install file on the OBIEE_SERVER box and also run the installer there? Or there a copy paste thing I do? What am I missing here?
I appreciate the help!
-Joe

I'm trying to install WebLogic as server for OBIEEI have not heard about OBIEE, but if your aim is to use weblogic as a server for hosting this then you definitely need to install weblogic on the physical box where OBIEE will be hosted.
I associated BIServer with the OBIEE_SERVER machine.
I went to start the BIServer and I got the following error:
"For server BIServer, the Node Manager associated with machine OBIEE_SERVER is not reachable."Again not sure what is BIServer but I think you are missing a number of steps here. If you are trying to run BIServer
on weblogic you definitely need to install WLS on this physical machine [OBIEE_SERVER]. You will then need to configure a node manager on OBIEE_SERVER machine to enable it to start remotely from the admin machine.
Link for Node Manager. Since you are in Linux you can try setting up script based node manager.
http://download.oracle.com/docs/cd/E13222_01/wls/docs92/server_start/nodemgr.html#wp1099239

ALSB2.6/WLS9.2 Node Manager Monitoring Issue

Hi,
We have configured a java based node manager on weblogic 9.2 platform, using our customized scripts, for ALSB domain.
Node manager uses a customized server startup script to start the servers. This start-up script does some custom tasks and then starts the weblogic managed server with java options specified in this start-up script.
We are able to start / stop the managed servers from admin console but the node manager is not able to monitor the health of the managed servers. Node manager's log file also shows that server is in failed state whereas the managed servers log files show running state.
Can you please help how can we make the java based node manager (which uses a script to start the server) to monitor/check health of the managed server.
Also we've noticed a change in the way parameters are passed to server start-up script (which is called by the node manager ) between WLS8.1 and 9.2.
In 8.1 arguments were passed to the script in$1,$2,$3 and $4 ($4 pointing to the pid file of managed server). However these arguments are no longer available in 9.2.

I think when nodemanager.properties runs at the start of node manager, there is one line in it, i.e.
StartScriptName=startWeblogic.cmd
now there are few lines in the startWeblogic.cmd, those are:
@REM Set IP Mask.
if NOT "%NetMask%"=="" (
set IPMASK=%NetMask%
) else (
set IPMASK=
@REM Perform IP Migration if SERVER_IP is set by node manager.
if NOT "%SERVER_IP%"=="" (
call "%WL_HOME%\common\bin\wlsifconfig.cmd" -addif "%IFNAME%" "%SERVER_IP%" "%IPMASK%"
now I think that's why everytime the node manager tries to start the server the IP address is changed or masked to different one in startup.properties file, what I have mentioned in previous post...and what is indicated in log file...
any guru...plz

Start managed server with the node manager???

Hello,
I have one admin server and one managed server for my domain. Can i Start the amdin and managed server from the node manager and not just the admin server. Currently with the install of node manager only the admin server starts and not the managed server. I need to log in to the admin server to start my managed server which is very cumbersome.
Edited by: user9021545 on Mar 11, 2011 1:58 PM

hi,
In order to start Managed server from the console itself,then u have to enroll your node manager.
Step1). Start the AdminServer...using startScript "startWebLogic.sh/cmd"
step2)start ur node manager using stat node manager scripts,
step3)Login to adminConsole and then see the NodeManager is Reachable or not
If reachanble
then run "setWLSEnv.sh/cmd".
Then run wlst.sh file
Step3)connect('weblogic','weblogic','t3://localhost:7001')
Step4)nmEnroll('C:/bea103/user_projects/domains/7001_Domain','C:/bea103/wlserver_10.3/common/nodemanager')
After this u will be able to start and stop managed server from console itself
Note:even after dis your managed server doesnt starts then please check under <DOMAIN_ROOT>\servers\<SERVER_NAME> whether you have a folder name as security and inside it whether you have file called as boot.properties
if no then follow these below steps
Please create a directory "security" inside your
<DOMAIN_ROOT>\servers\<SERVER_NAME>
inside "<DOMAIN_ROOT>\servers\<SERVER_NAME>\security" ( Example Location:
WLS103/user_projects/domains/Test_Domain/servers/AdminServer/security )
directory create a file with exactly same name as "boot.properties" and with
following 2 lines inside it:
username=weblogic
password=weblogic
NOTE: Here "weblogic" is the Administrator username & password...
if you want dont want to run Admin server from foreground then try runnin it as a background process
This link wi; b helpful for that
http://middlewareforum.com/weblogic/?cat=23
any further help in this den do let me know
Regards
Fabian

Cannot set node manager password in weblogic console (10.3.3)

Hi,
I've been trying to configure a custom AD authenticator in the weblogic console for 10.3.3 and get Invalid pad byte on the node manager password when it tries to activate my changes. After trying several things suggested in other forums, I went into the weblogic console, under domains->security, changed the node manager user and password, and still get an Invalid pad byte on the password I entered! Any suggestions?
Thanks!
Julia Allen

Well that didn't work :( . I went into config.xml and changed <node-manager-username> and <node-manager-password-encrypted> to clear text, restarted, went into the console and tried to change the node manager username and password in Domains-->Security-->Advanced and still got invalid pad byte on the password I entered when I saved. Interestingly enough when the page came up the value for node-manager-username was the old value... not the one I changed in config.xml before I started the server. I verified that I changed the correct config.xml. When I looked at it after restart, the values had not been changed to encrypted values. And, yes, production mode is false. So I'm guessing a cache problem somewhere? Or a corruption in one of the other security files that I cannot determine?
One thing I did try was to reset the admin password following steps I found in other places (e.g. renaming ldift/initialized/properties files, running java weblogic.security.utils.AdminAccount) but nothing. Still the same problem. Also note that the server starts just fine... the console just barfs with this invalid pad byte error whenever I try and configure that AD provider or actually try and change the node user name and password.
Any other ideas? Your help is greatly appreciated!
Julia

After I do Node Manager nmConnect(...) I can no longer start a managed svr

I am running 10.3.2 on both windown and redhat
I have a domain with AdminServer and one managed server.
I can start AdminServer from the shell with startWebLogic.cmd
I can log into the web based 'console'
I can go to Environment -> Servers and start/shutdown the managed server.
If I use WLST and...
connect('user', 'pass', 't3://localhost:7001')
nmEnroll('fullPathToDomainDir')
I am still OK, I can still use the console to start and stop the managed server and I can so WLST start('managedSever')
I stop running the AdminServer from the shell - it is no longer running.
Now, I want to start the AdminServer with Node Manager, so (with WLST)...
nmConnect('user','pass','localhost','5556','domainName)
I get the error...
'Cannot connect to Node Manager: Access to domain 'domainname' for user 'user' denied
As suggested elsewhere, I go and edit nm_password.properties and change the hashed user/pass to
username=user
password=pass
I go back to WLST, now nmConnect works!!!
I can then do..
nmStart('AdminServer')
The AdminServer start, I go to the web console, and I try to start the managed server, I get...
'For server managedServer, the Node Manager associated with machine machineName is not reachable'
So it looks like the re-hashed password from nmConnect(...), while it works for WLST, it no longer works for the console.
Bottom line, I can get either get the console to start the managed server (if I do WLST connect(...) + nmEnroll(...)) or I can get WLST to start the servers (if I reset nm_password.properties and do nmConnect), but I can not get WLST to start the AdminServer and have the AdminSever console start the managed server.
I know for a fact that others have run into this, but in the threads I found, the thread stops before a soluton is posted.
Anyone know about this issue?
Regards

There are a couple of things you need to keep in mind:
-- There is one and only one valid username / passwrod for connecting to the node manager.
-- When you use nmEnroll, the username / password placed in the nm_password.properties file is taken from the admin server, hashed an placed in the nm_password file.
-- When you edit the nm_password.properties, and place the line "enter in username=user / password='pass',", in clear text, you have over ridden the U/P \thje admin server knows.
That is why you have a situation, oen of the two works but never both at the same time.
What you need to do is:
-- Log in to the WL Console.
-- in the Domain Tree, click at teh root level, which is the domain name.
-- Under the security tab, under advanced there is a node manager username / password.
-- Click on "Lock and edit", and change those to match what you are using through WLST.
-- Activate changes.
-- Just to make sure all is well, use nmEnroll to re enroll the node manager again , this way the nm_password.properties file will be set up correctly.
Now, both the WL Console should be able to talk to the node manager as well as nmConnect('user','pass','localhost','5556','domainName) would succeed.
Also, if you want to use node manager to start /servers, make sure you edit the file ( $WL_HOME/common/nodemanager/nodemanager.properties) and change the line
StartScriptEnabled=false
to
StartScriptEnabled=true
This will ensure all the proper class paths and properties etc needed are set up when using the node manager.
You would need to bounce the node manager for this change t take effect
Hope this helps.

WL 7.0 Node Manager - auto restart after system boot fails

I have configured node manager to automatically restart managed servers on my Solaris
2.8 host. When I kill the managed server from the command line, node manager automatically
restarts the managed server just as I would expect. But when I reboot the system,
node manager does not restart the managed server. I started the managed server before
the reboot using node manager/admin console as the documentation states. I see nothing
in the logs that indicate any attempt by node manager to start the previously running
managed server, although node manager itself starts up OK after the reboot. Any
suggestions on how to get this working?
Here is my node manager config.xml contents:
#Saved configuration for webdevDomain_apps
#Sun Jul 14 23:12:08 EDT 2002
serverListenProtocol=http
serverListenHost=198.242.140.49
serverHealthCheckIntervalSeconds=180
TimeStamp=1026702630479
javaHome=/usr/jdk1.3
serverRestartDelaySeconds=60
command=online
ts2={3DES}QQj3hoYUACzB1csbrEth8Q\=\=
ts1=system
serverHealthCheckTimeoutSeconds=60
weblogic.Name=apps
java.security.policy=/home/weblogic/weblogic700/server/lib/weblogic.policy
RootDirectory=/home/webdev/weblogic700
serverStartArgs=-Xms64m -Xmx256m
- John

you are 100% right.
NM restarts MS's incase of
i) server crashes
ii) system reboots.
Kumar
John wrote:
Actually we were given a solution to this problem:
The node manager (NM) should restart managed servers (MS) upon reboot of the box.
There is one exception however - this can't happen if the MS is part of a cluster
and it hosts a migratable service. Currently there are two migratable services
- JMS and JTA. You need to have explicitly setup the migratable services for the
MS via the console.
Assuming you are not using these as migratable services you must explicitly set
a flag to indicate the MS is not hosting migratable services. This can be done
via the command line. It only needs to be done once and it will update the config.xml
file. The command line is:
java weblogic.Admin -url t3://adminhost:port -username <user> -password <password>
SET -mbean <domain>:Name=<MS name>,Type=Server -property HostsMigratableServices
false
You should see a simple reply of 'OK'.
Once this is done your should be able to reboot the box and after restart the
managed server should be running.
It worked!
Kumar Allamraju <[email protected]> wrote:
NM will only restart managed server's if they are abruptly killed.
We have no mechanism inplace to start managed server's upon system reboot.
May be you should write some startup scripts to start WLS upon system
start.
Kumar
John wrote:
I have configured node manager to automatically restart managed servers
on my Solaris
2.8 host. When I kill the managed server from the command line, node
manager automatically
restarts the managed server just as I would expect. But when I reboot
the system,
node manager does not restart the managed server. I started the managed
server before
the reboot using node manager/admin console as the documentation states.
I see nothing
in the logs that indicate any attempt by node manager to start the
previously running
managed server, although node manager itself starts up OK after the
reboot. Any
suggestions on how to get this working?
- John

Setting up the Script Based Node Manager

Similar Messages

Maybe you are looking for