faq_AdminGUIandCLI

Frequently asked questions about Administration GUI or CLI

Web Server 7.0 Administration Terminology

1. Where are the configurations stored on the file system?

All the configurations corresponding to the instances managed by the admin-server are stored in admin-server/config-store directory.

Web Server 7 Administration GUI

 1. How do I access the Web Server 7.0 administration console?

The admin console can be accessed by pointing a browser tohttps://<admin-host>:<admin-ssl-port>. The admin SSL port defaults to 8989. This will bring you to a login page wherein you must type in the administrator user name and password that you specified when you installed the admin server.

2. What is the certificate warning that I see when I first access the console?

This warning shows up because the admin server has a self-signed certificate installed on it and the browser prompts you to check whether you trust the certificate.

3. I seem to get lost in the GUI. How do I figure out where I am? Are there any navigational aids that I can use?

The best indication of where you are in the GUI is the bread-crumbs that you see above the tabs, they tell you where you are within the object hierarchy. If you totally get lost, you can always click the Home button in the masthead that will take you to the Common Tasks page.

4. How do I enable or configure SSL using Web Server 7 Administration GUI

Here is a great blog describing this topic.  Feel free to post to our forum if you need any further help 

5. When admin. nodes are not running, then Web Server 7 Administration Server GUI appears "hung" ?

Try to shorten the timeout values below if needed, e.g.  in the <install_dir>/admin-server/config/server.xml:   

<jvm-options>-Dsun.net.client.defaultConnectTimeout=60000</jvm-options>  
<jvm-options>-Dsun.net.client.defaultReadTimeout=120000</jvm-options> 

See more at http://java.sun.com/j2se/1.5.0/docs/guide/deployment/deployment-guide/jcp.html  

Web Server 7 Administration CLI

1. Where can I find the CLI for Web Server 7.0 administration?

The WS 7.0 administration CLI is located at <install-root>/bin/wadm. To use the CLI, you need to know:

admin server hostname (defaults to localhost)
SSL port for the admin server (defaults to 8989)
admin server user name (defaults to admin)
admin server password
Note that the Administration Server needs to be running in order to use the CLI. You can start the admin server by running <install-root>/admin-server/bin/startserv.

2. How can I get help on a CLI command?

Help can be accessed at any time using the --help option. wadm --help gives the overall top-level help. wadm <sub-command> --help gives the command specific help.

3. Why does invoking the CLI using wadm --user=admin --host=myhost sub-command just take me into a command shell instead of executing the command?

This is because the sub-command needs to precede all the options including common options like --user and --host; else the CLI will assume that you are invoking it in the shell mode and take you to a shell prompt.

4. Do I need to provide user, port, host, passwordfile options even in the shell mode?

The connection to the admin server is made while starting the shell, hence these options are not required for sub-command invoked within the shell. Even if they are provided, they will be ignored

Sometimes when a user connects to the server using the CLI for the first time, there is a cryptic message about a certificate and whether I trust the certificate. What should the user do?
The CLI connects to the SSL port of the admin server by default. The admin server has a self-signed certificate installed on it and the message that you see when you connect to the server for the first time is whether you trust the certificate. Make sure you are connecting to the right server and port and say yes to trust the certificate. For subsequent invocations, this message is no longer displayed.

5. Does the CLI have an interactive mode of execution?

If you invoke the CLI without the --commandsfile option, it enters the interactive mode or the "shell" mode. The shell mode has tab-completion for commands and you can use arrow keys to go through the command history and to edit previous command invocations.

6. Can I put the commands into a file and execute the CLI in the "batch" mode?

Yes, you can put the CLI commands into a file and invoke the CLI using the --commandsfile <filename> option. In fact, the CLI embeds a full-featured jacl scripting engine (jacl is a Tcl engine written in java) and you can take advantage of a full scripting environment interspersed with command execution.
e.g.: wadm --user admin --port 8989 --host hydrogen.india --passwordfile=~/wadmpwdfile --commandsfile=/space/wadmscr
Note that for complete jacl integration, you need to specify:
set wadm_script true
either in the script or in the rcfile.

7. How can I add some initialization to the command line that will be used on every invocation?

The CLI reads a file named .wadmrc residing in the user's home directory ($HOME) and treats it as a startup file. Note that any option can be supplied to the commands by setting it within the shell or the .wadmrc file. This avoids typing the common options for every command. The variable name equivalent for an option is derived by prefixing wadm_ to the option name and replacing all '-' in the option name with '_'.
e.g.:
option: user
shell variable: wadm_user
option: replace-algo
shell variable: wadm_replace_algo

The rcfile(.wadmrc) contents to set the above options will be as follows:
set wadm_user admin
set wadm_replace_algo lifo

8. Can I change the name/location of this startup file?

Yes, you can. You can pass in a startup file to the CLI using the --rcfile option.

9. Can I configure the CLI not to prompt for a password each time?

Yes, you can provide the admin-user's password in a file which must contain the password in the following format:
wadm_password=<password>

Note that some of the CLI commands also accept passwords, e.g. create-user. User can specify these passwords in the same password file.
For example, bind-password can be supplied through the password file as:
wadm_bind_password=<bind-password>
You can turn off password prompting altogether by specifying the --noprompt option on the CLI.

For commands like create-user have password options, it is inconvenient to provide a password file each time. Is there any other way to provide passwords?
In both the shell mode and in the --commandsfile mode, passwords can be specified just like regular options.
e.g.:
$wadm --user=admin --port 8989 --passwordfile=~/.wadmpwdfile
wadm>create-user --config=test --authdb=db1 --user-password=abc123 user1
wadm>create-user --config=test --authdb=db1 --user-password=xyz123 user2
wadm>exit
$

10. How can I see what the CLI is sending to the backend?

The --echo option prints the whole command with all the options and their values supplied by the user, read from the .wadmrc file or assumed as default by wadm. This is NOT a documented interface, but support folks can also modify the <install_root>/bin/wadm script and add a -DDebug jvm option to the java invocation to get verbose debugging information.

11. Are there any other common CLI options that I should be aware of?

The -verbose option on the CLI prints out informative messages when there are no values returned from the backend, when a command succeeds, etc. It also prints out column header names in the case of list commands. The -all option on the list commands prints out extra columns of information.
e.g.:
$wadm list-instances --config=test
hydrogen

$wadm list-instances --verbose --config=test
node-name
---------
hydrogen

$wadm list-instances --all --verbose --config=test
node-name instance-status
-----------------------
hydrogen running

12. The sub-commands and option names are too long, is there a way to shorten them?

Unfortunately, there are no short forms for the sub-commands. For each of the options, there is a corresponding single letter short-option which you can use.
$wadm command-name --help
prints a man page which contains short option name mappings for long options.

13. Does the CLI support exit codes?

In general, the CLI returns 0 for success and a non-zero value indicates an error. We are currently in the process of adding exit codes and will be publishing these shortly.

14. What is Deployment Pending ?

When I make any changes to a configuration from the GUI, I see a Deployment Pending message in the masthead. What does it indicate?
The "Deployment Pending" message indicates that there are changes to the configuration that are saved into the admin-server's config-store. You will now need to click on Deploy in order to propogate these changes to the instances.

15. What is the difference between the Save and Deploy buttons that seem to show up on every page within a configuration?

If you click on Save, then the changes are saved only into the config-store. Only when you click Deploy will the changes be propogated to the instances.

16. Can I make manual edits to my configuration files? Will these be reflected in the CLI and GUI?

If you need to make manual changes to the config files, it is recommended that you make these changes within the web server instance and then pull these changes back into the admin-server using the administration GUI or pull-config CLI.

17. I made some changes to the server configuration using the CLI. What command should I use to "apply" these changes to a running server instance?

You need to use the deploy-config command to apply changes to a running instance.

18. When I try to invoke deploy-config, it fails saying that 1 or more instances have been changed, what does this mean?

This means that you have probably made some manual changes to the configuration files. You now have 2 options:

• discard the manual changes by invoking deploy-config with the --force option
• pull the manual changes into the config-store using pull-config and deploy the same to all the instances

19. Sometimes a deploy-config fails saying that a server restart is needed, why does it do this?

When a deploy-config is invoked, the configuration changes in the config-store are propogated to the instances and the running servers are reconfigured so that the config changes are picked up. But certain changes cannot be applied dynamically and need a server restart, you will hence need to run a restart-instance command in order to restart the server instances.

20. How do I tell the server not to reconfigure the instances after a deploy?

By using a --reconfig=false option on deploy-config.

21. What do deploy-config, create-instance, delete-instance succeed even when a particular node associated with a config is down?

This is so that the failure of a single node does not impact the operation on the other nodes. There is a synchronization module that ensures that when the node that is down comes up, it will be sync-ed up with the latest data from the config-store.

22. How to configure SSL using Web Server 7 Admin CLI

Here is a nice blog on how to do this t