|
|
This chapter contains the procedures and ancillary information necessary for successfully using WLS-VE. It shows how to create, start, and stop virtualized servers in a WebLogic Server domain, start and run this product, and how to use its features. It also describes conditions that must be met before you attempt to use the product. In some cases, WLS-VE procedures are identical to the same procedures in non-virtualized WebLogic Server. In those cases, the chapter will provide a brief description of the functionality and links to the relevant WebLogic Server documentation.
This chapter contains information on these subjects:
This section describes requirements for running WLS-VE successfully. You will probably have completed these sorts of tasks as a general practice; however, verifying that they are already met before you begin will reduce problems later on.
Before attempting to run WLS-VE, ensure the following:
| Note: | Please review important security information about the NFS servers in NFS Security Measures. |
If you have experience using non-virtualized WebLogic Server, you might want to use some of the configuration techniques common to that product. If so, you need to be aware that some of them will not work with WLS-VE.
boot.properties and SerializedSystemIni.dat files. You cannot specify credentials at an interactive prompt, because WLS-VE does not offer one.<domain>/servers/<myserver>/logs/... The other one is the LiquidVM Console Log. The Console Log is normally stored in the current working directory as seen from inside the WLS-VE instance. By default, that is /domain (that is, your domain directory on the NFS server). The default name of the file is <vmname>.Use your favorite tools (for example, tail) to monitor these log files.startWLSVE.sh startup script that was created will include that Administration Server as a parameter.
To create a domain and WLS-VE servers ("virtualized servers"), use this procedure.
| Note: | This procedure assumes the following: |
The Configuration Wizard launches.
startWLSVE_Admin.sh startup script that was created will include that Administration Server as a parameter.| Note: | Unlike with non-virtualized WebLogic Server, you cannot reuse an IP address for multiple servers |
Be sure to note where you create your domain (assuming you can put this domain on the NFS server) and what you named your Managed Servers.
| Note: | If you are going to boot the Administration Server on a virtual machine, Start Admin Server on the Creating a Domain window should remain unchecked. |
<DOMAIN-DIR>/bin/and locate the file startWLSVE.sh (or .cmd on Windows) as described here:
The bin/startWebLogic.cmd and startWebLogic.sh files are not valid for WLS-VE.
startWLSVE_Admin.sh (or .cmd) and startWLSVE_Managed_nn.sh (or .cmd) by setting the properties required for your implementation. At the minimum, you must set the following values:IP_ADDRESS to a valid IP address. This must be an address that no one else will be attempting to use when you are trying to run your machine. Otherwise, you will receive a "Configured IP [...] in use by MAC" Error. You test an IP address for availability using ping; if there is no response to ping at that address, it is probably available. Check with your network administrator.DOMAIN_MOUNT to the NFS mount for this domain. Specify this property like this:DOMAIN_MOUNT=IPA.DD.RE.SS:/full/path/to/domain,uid=nnnnn,gid=nnnnn
where gid is the group identifier and uid the user identifier (see Configuring Group and User Names for an explanation of these identifiers); for example:
DOMAIN_MOUNT=131.18.110.61:/space1/servers/bea_home/user_projects/domains/product-1,uid=10016,gid=10000
BEA_HOME_MOUNT to the location of your <BEA_HOME> on your NFS server. Specify this property like this:BEA_HOME_MOUNT=IPA.DD.RE.SS:/full/path/to/BEA_HOME,uid=nnnnn,gid=nnnnn
where gid is the group identifier and uid the user identifier (see Configuring Group and User Names for an explanation of these identifiers); for example:
BEA_HOME_MOUNT=131.18.110.61:/space1/servers/bea_home,uid=10016,gid=10000
TMP_MOUNT to the location where the WLS-VE instance will store temporary files. Specify this property like this:TMP_MOUNT=IPA.DD.RE.SS:/full/path/to/temp_directory,uid=nnnnn,gid=nnnnn
where gid is the group identifier and uid the user identifier (see Configuring Group and User Names for an explanation of these identifiers); for example:
TMP_MOUNT=131.18.110.61:/space1/servers/server_temp,uid=10016,gid=10000
Additionally, for each Managed Server, verify that the ADMIN_URL is pointing to the IP address of the Administration Server (for the Administration Server, ADMIN_URL must be blank).
| Note: | The start-up scripts do not and should not include a path reference to the WebLogic Server classes. WLS-VE uses the WebLogic Server classes that are located in the WLS-VE ISO image. |
| Note: | To create a Managed Server in your WLS-VE domain, it is not enough to create a start-up script; you must also create the Managed Server instance in the domain, as described in step 4. |
<DOMAIN_DIR>/bin/directory and entering the start-up command; for example:
startWLSVE_Admin.sh (or .cmd on Windows)
startWLSVE_Managed_nn.sh (or .cmd on Windows)
The server output will appear at <DOMAIN_DIR>/WLS-<servername>/log, unless you have specified a different location using the LVM_CONSOLE_LOG environment variable.
| Note: | The message JVM is booted does not mean that your server is fully running. It means that your virtual machine has been started and that your server startup process has been initiated. |
In addition to the required configuration properties for the Administration Server and Managed Servers, you can configure other aspects of your VM by setting any of the properties shown in the start-up script. Please refer to Additional Configuration Tasks for details.
When LiquidVM boots, it has to mount the BEA_HOME and DOMAIN directories, which are located on an NFS server. Any time you mount an NFS share, you provide a user and group name. When you attempt to read or write below the NFS share, the NFS server compares the user/group ID (uid and gid, respectively) with the access control list for the file or directory that you requested. You have to ensure that the user and group ID that you supply in the mount command has sufficient privileges to read and write in the domain directory. Note that if the domain directory is one or more directory levels below the share, the uid and gid must be able to read all of the directories from the share down to the domain directory. It must be able to read and write in the domain directory itself.
To ensure that read and write permissions are set up properly, run the Domain Configuration Wizard from the same OS user account that you plan to use to mount shares.
/myShare directory as an NFS share. On the machine on which you ran the WLS-VE installer, you log in as user "pat" and run the Domain Configuration Wizard to create a domain in /myShare/user_domains/base_domainuid and gid for "pat." To do this, use the UNIX command id. Your uid(s) and gid(s) will be displayed. For this example, let's say that they're 1112 and 1115./myShare/user_domains/base_domain/bin/startWLSVE_Admin.sh startup script, for the DOMAIN_MOUNT variable, specify: set DOMAIN_MOUNT=myNFSServer.mycompany.com:/myShare/user_domains/base_domain,uid=1112,gid=1115
Pat's uid and gid must have read privileges for myShare and user_domains. They must have read and write privileges for base_domain and everything below it.
When you install WLS-VE, the installer generates a DemoIdentity keystore that includes a certificate for the host that runs the installer. This is not the host from which you will run the product. The installer then stores this certificate on the ISO. If SSL is used, this certificate is the default if no other certificate is specified.
When you launch WLS-VE, it will use a different IP address than the installation machine. WebLogic Server will generate a warning that the hostname is not the same as the host for the certificate. You can ignore this warning and the machine will run normally.
Generally, upgrading and promoting virtualized domains requires the same steps used for upgrading and promoting non-virtualized domains. The main steps in this process are:
In this step, you need to inventory the application environment, verify supported configuration information, review the compatibility information, and create an upgrade plan
In this step, you undeploy any deployed applications, shut down all servers in the application environment, back up the application environment, install any required BEA products, prepare the remote Managed Server domain directories, and set up the environment
For detailed instructions on these steps, please see Upgrading WebLogic Application Environments.
Due to its virtualized nature, when you upgrade WLS-VE, depending upon your required upgrade scenario, you will have to modify the standard upgrade procedure to address important virtualization issues. Those are described below.
When you upgrade an application from a non-virtualized implementation of WebLogic Server 9.2 to WLS-VE 9.2, you will need to make some modifications to the application code to ensure successful operation.
<DOMAIN-DIR>/bin/startWLSVE.sh or .cmd file, and edit the start file to work with your domain, as described in Creating WLS-VE Domains. java to java_esx.localhost as the host name, you must change the host name to the LiquidVM IP address.| Note: | If your application uses Pointbase, you will need to change localhost to the IP address of Pointbase. |
connect() command, you will need to change it so that the hostname can be passed in.You cannot migrate an application directly from an earlier version of WebLogic Server (for example, version 8.1) directly to WLS-VE 9.2. Instead, you need to follow the upgrade procedures outlined in Upgrading WebLogic Application Environments.
Moving a WLS-VE domain into a production environment follows the same procedure used to move any development domain into production. The critical concerns here relate to added security and performance requirements in a production environment. See Securing a Production Environment for suggestions on securing your production environment. You also need to understand the NFS security issues described in NFS Security Measures. As with any migration, be sure to migrate the VM and any files stored on NFS server at the same time.
As defined earlier, WLS-VE is WebLogic Server plus LiquidVM, which is BEA JRockit on BEA's BareMetal technology. When you start LiquidVM, it will simultaneously boot the associated server. You can start a WLS-VE instance either from the command line or from VMware VirtualCenter running on the ESX host.
| Note: | You cannot start a WLS-VE server instance from the WebLogic Server Administration Console, nor can you use the Administration Console to resume or suspend a WLS-VE server instance. You can and should use the Administration Console to shut down a WLS-VE server instance. |
To start WLS-VE from the command line:
<DOMAIN_DIR>/bin/.After you have started WLS-VE from the command line at least once, you can start it from the VMware VirtualCenter:
| Note: | Please see DemoIdentity Keystore Warning for important information about a WLS-VE startup warning. |
Stopping WLS-VE should be carefully considered as the ripple effects of an improper shutdown can cause unexpected results. Ideally, you should shut down WLS-VE only by using one of following methods:
| Note: | You should not shut down WLS-VE by using the following methods: |
Since WLS-VE contains both a JVM and a virtualized WebLogic Server instance, you can configure both devices by using the same configuration flags used by their non-virtualized editions. Usually, you can do this from the WebLogic Server administration console. Please refer to System Administration for BEA WebLogic Server 9.2 for complete information on how to:
The JVM should already be well-tuned for most WebLogic Server applications but you can configure and tune a machine's Java behavior by setting the necessary Java options in the start-up script for the domain in question. Simply enter the standard J2SE start-up options or BEA JRockit's non-standard -X and -XX options at the JAVA_OPTIONS= statement.
Listing 7-1 shows a snippet of the Administration Server start-up script, startWLSVE.cmd, with JAVA_OPTIONS= highlighted.
@ECHO OFF
SETLOCAL
.
.
.
set PRE_CLASSPATH=
set POST_CLASSPATH=set JAVA_OPTIONS=set JAVA_PROPERTIES=
For example, suppose you want to start the machine so that LiquidVM uses a garbage collector (that is, a memory management system) optimized for application throughput. You would do this by setting JAVA_OPTIONS thusly:
JAVA_OPTIONS="-xgcprio:throughput"
You can string together as many valid options as you need; however, you must place them within quotation marks and separate them with a single space. For example, the following code:
JAVA_OPTIONS="-xgcprio:throughput -xgcreport -Xss:512k"
in VirtualCenter"-xgcreport"). See the BEA JRockit Reference Manual for a list of valid LiquidVM start-up options and instructions for using them. For LiquidVM tuning and configuration guidelines, see Tuning Basics and Memory Management Basics in the BEA JRockit Diagnostics Guide.
In addition to JVM behavior, you can also configure WLS-VE behavior by setting other parameters in the start-up script. The properties you can set are listed in Table 7-4.
Deploy applications on WLS-VE the same way you deploy them on non-virtuali.zed WebLogic Server. Application deployment generally involves the following tasks:
weblogic.Deployer These tasks are detailed in Deploying Applications on BEA WebLogic Server 9.2.
|