Chapter 4. Managing Clients

Go here for info on the current status of SCO Linux products
Table of Contents
4.1. Setting up Polling
4.1.1. Creating a Polling Policy
4.1.2. Creating a Polling Action
4.2. Configuring Alerts
4.2.1. Using the Setup Logging Wizard
4.2.2. Manually Creating a Gateway Policy
4.3. Performing Inventories
4.3.1. Performing a Software Inventory
4.3.1.1. Using the Run Software Inventory Wizard
4.3.1.2. Manually Running a Software Inventory
4.3.2. Performing a Hardware Inventory
4.3.2.1. Using the Run Hardware Inventory Wizard
4.3.2.2. Manually Running a Hardware Inventory
4.3.3. Setting the Number of Inventories to Keep
4.4. Working with Computer Groups
4.4.1. Creating a Computer Group from Inventory Results
4.4.2. Manually Creating a Computer Group
4.5. Monitoring System Health
4.5.1. Creating a Health Policy
4.5.2. Creating a Health Action
4.5.3. Creating a Custom Action
4.6. Managing Printers
4.6.1. Creating a Printer Policy
4.6.2. Creating a Printer Action

Caldera Volution Manager provides a number of built-in tools that allow you to manage your clients, either individually or as a group. For example, you could inventory all clients containing graphics cards made by a particular vendor, create a computer group based on those computers, and then install an updated driver package that you downloaded using Caldera Volution Online on only those systems.

In addition to the actions you can implement directly, you can also create your own scripts to perform actions not directly supported by VM. For example, you might create a script to kill runaway processes or remove temporary files when free disk space is low.

Two types of special objects are used to accomplish these tasks: actions and policies. Actions are scripts that implement a task at a particular time. Policies govern the scope of the action. Linking both to a specified computer object or container executes the action at the scheduled time.

This topic includes:

These topics assume knowledge of linking, searching, and managing objects.

4.1. Setting up Polling

By default, Caldera Volution Manager uses the Distributed Events Notification Service (DENS) to inform clients of directory service changes that affect their configuration. In some cases, clients cannot access DENS, either due to a network firewall or their connection status (some systems are not always be connected to an intranet). These systems must instead be set up to poll the directory service to obtain updates.

To set up a system to poll the directory service:

  1. Create a new DENS policy for the computer(s) you want to poll.

  2. Create a Poll for directory changes action.

  3. Link both the new DENS policy and the new action to the computer(s) you want to poll.

The following describes these steps in detail.

4.1.1. Creating a Polling Policy

To create a policy that enables polling:

  1. Open the policies container.

  2. Select Create an Object.

    1. Select DENS Policy from the drop-down list.

    2. Enter a unique name for the polling policy (such as Poller).

    3. Click Create.

  3. Select the new policy from the Contents pane.

  4. Select Edit the Data.

  5. Uncheck Use DENS if it is selected, then click OK.

  6. Select Add or Remove Links and Link the policy to the desired computers or containers.

  7. Click Save Changes.

The computers you linked to no longer use DENS. You must now configure a polling action as described in Section 4.1.2.

4.1.2. Creating a Polling Action

To create a polling action for computers that do not use DENS:

  1. Open the actions container.

  2. Select Create an Object.

    1. Select Action from the drop-down list.

    2. Enter a unique name for the polling action (such as Poller).

    3. Click Create.

  3. Select the new action from the Contents pane.

  4. Select Edit the Action.

  5. Select Poll for directory changes as the Action to Run and click OK.

  6. Select Edit the Schedule and define a schedule (see Section 2.4.1.3).

  7. Link the action (see Section 2.4.2) to the same computers or containers to which you linked the poller policy.

  8. Click Save Changes.

4.2. Configuring Alerts

Caldera Volution Manager can be configured to issue alerts that are triggered when there are changes to the directory structure or when health monitor thresholds are exceeded. These alerts can be logged either to an e-mail address using SMTP or to an SNMP console. You can choose the types of alerts you want to monitor, and even have different alert types and logging methods for different computers or computer groups.

Logging methods and alert types are configured using a gateway policy. A sample gateway policy is included in the policies directory; however, you should create at least one new policy to manage alerts on your network.

Table 4-1. Alert Types

Alert TypesExplanation
emergencyLog message if the system becomes unusable
alertLog message if immediate action is required (use for health monitoring)
criticalLog message if a critical condition occurs
errorLog messages if an error condition occurs (used by volutiond)
warningLog warning condition messages (use for health monitoring)
noticeLog normal but significant conditions (use for health monitoring)
informationLog informational messages (used by volutiond)
debugLog debug-level messages (used by volutiond)

4.2.1. Using the Setup Logging Wizard

The Setup Logging wizard guides you through the process of creating an gateway policy to set up logging.

  1. Click Wizards->Setup Logging.

  2. Follow the directions presented on the screens.

The wizard asks you to supply:

  • a unique name for the gateway policy you are creating

  • whether you want to log messages to an SNMP console or to an e-mail address (SMTP)

  • the fully qualified host name or IP address of the SNMP console address and the Community (default is Public), or the fully qualified host name of the SMTP server and the e-mail address you want to send messages to

  • the Alert Types you want to monitor

  • the individual computers, computer groups, or containers that you want to link the gateway policy to

4.2.2. Manually Creating a Gateway Policy

To manually create a new gateway policy:

  1. Open the policies container.

  2. Select Create an Object.

    1. Select Gateway Policy from the drop-down list.

    2. Enter a name for the gateway policy (such as Standard Alerts).

    3. Click Create.

  3. Select the new policy from the Contents pane.

  4. Select Edit the Gateways.

  5. Click Add to add a gateway.

    • To add an SMTP (e-mail) gateway, select SMTP from the drop-down list and click Add. Then enter the mail server address (a fully qualified host name; for example, mailserver.yourcompany.com) and click OK.

    • To add an SNMP (SNMP console) gateway, select SNMP from the drop-down list and click Add. Then enter the SNMP server address (the fully qualified host name or IP address) and click OK.

  6. Select Edit the Gateways again, and click on the name of the gateway for which you want to set alert levels.

    • For SMTP, enter the e-mail address(es) of the accounts to receive the alerts, then click Add.

    • For SNMP, enter the community in the Community field. If no Community is entered, the default is Public. You can enter any string you want. Machines can be configured to only accept instructions from a system that passes a particular Community string, so Community can be viewed as a kind of password, but it is not encrypted.

  7. Select the Alert Types (Alert Log, Error Log, and so on) that you want to be notified about.

  8. Repeat steps 6 and 7 for each gateway you configure and when done, click OK.

  9. Select Add or Remove Links and link the policy to the desired computer or computer group.

  10. Click Save Changes.

4.3. Performing Inventories

Caldera Volution Manager enables you to perform software or hardware inventories of individual computers or computer groups.

4.3.1. Performing a Software Inventory

A software inventory lists the packages (RPM, pkg, or custom) currently installed on a client system. You can use the Run Software Inventory wizard to create an action to run a software inventory, or you can manually run a software inventory.

You can customize your software inventories:

  • by creating a different schedule for the inventory action (see Section 2.4.1.3)

  • by adding constraints to the action (see Section 2.4.1.2)

  • by linking to more than one computer, or to a computer group (see Section 2.4.2)

  • by defining the number of inventory scans to keep (see Section 4.3.3)

4.3.1.1. Using the Run Software Inventory Wizard

The Run Software Inventory wizard guides you through the process of creating an action to run a software inventory.

  1. Click Wizards->Run Software Inventory.

  2. Follow the directions presented on the screens.

The wizard asks you to supply:

  • a unique name for the software inventory action you are creating

  • the computer, computer group or container that contains software you want to inventory

  • when you want the inventory to begin

4.3.1.2. Manually Running a Software Inventory

To manually run a software inventory (in this example, the action runs once on the computers in the computers container):

  1. Select the Run Software Inventory action in the Actions container.

  2. Select Edit the Schedule, select a schedule type, and click OK.

    By default, the inventory is scheduled to run immediately (when the client starts or when the action changes).

  3. Select Add or Remove Links.

  4. Click Add.

  5. Enter computers in the Search expression field and click Search.

  6. Select the computers you want to inventory from the Search Results and click Link.

  7. Click Save Changes.

A software inventory appears in the directory tree beneath each inventoried computer.

4.3.2. Performing a Hardware Inventory

A hardware inventory lists packages currently installed on a client system. You can use the Run Hardware Inventory wizard to create an action to run a hardware inventory, or you can perform the procedure manually.

You can customize your hardware inventories in any of these ways:

  • by creating a different schedule for the inventory action (see Section 2.4.1.3)

  • by adding constraints to the action (see Section 2.4.1.2)

  • by linking to more than one computer, or to a computer group (see Section 2.4.2)

  • by defining the number of inventory scans to keep (seeSection 4.3.3)

4.3.2.1. Using the Run Hardware Inventory Wizard

The Run Hardware Inventory wizard guides you through the process of creating an action to run a hardware inventory.

  1. Click Wizards->Run Hardware Inventory.

  2. Follow the directions presented on the screens.

The wizard asks you to supply:

  • a unique name for the hardware inventory action you are creating

  • the computer, computer group or container that contains software you want to inventory

  • when you want the inventory to begin

4.3.2.2. Manually Running a Hardware Inventory

To manually run a hardware inventory (in this example, this action runs once on the computers in the computers container):

  1. Select the Hardware Inventory action in the Actions container.

  2. Select Edit the Schedule, select a schedule type, and click OK.

    By default, the inventory is scheduled to run immediately (when the client starts or when the action changes).

  3. Select Add or Remove Links.

  4. Click Add.

  5. Enter computers in the Search expression field and click Search.

  6. Select the computers you want to inventory from the Search Results and click Link.

  7. Click Save Changes.

A hardware inventory appears in the directory tree beneath each inventoried computer.

4.3.3. Setting the Number of Inventories to Keep

The Inventory policies specify how many inventories to keep. For example, if the number is 7 (default), then the results of the seven most recent inventory scans are kept. Inventory scan objects are stored in the scans containers under each computer object. A scan object can be thought of as a kind of differential file comparison: it has entries for everything added or removed since the last scan.

To change the number of scans to keep:

  1. Open the policies container.

  2. Select Create an Object.

    1. Select Inventory Policy from the drop-down list.

    2. Enter a unique name for the policy (such as Save3HW).

    3. Click Create.

  3. Select the new policy from the Contents pane.

  4. Select Edit the Scans.

  5. Enter the number of scans to keep and click OK.

  6. Link the policy to the desired computer(s) or computer group(s) to which you have linked the associated inventory action.

  7. Click Save Changes .

4.4. Working with Computer Groups

When you link actions, profiles, and policies to a computer group, all of the computers in the group are affected. If you have computers with similar characteristics or purpose, it is quicker and easier to work with a group than with each system individually. You can create additional groups based on:

An individual computer can be linked to more than one group. After a group is created, it can be populated by adding links to it from other computer objects. Computer groups can exist anywhere in the directory structure.

4.4.1. Creating a Computer Group from Inventory Results

When you perform a hardware or software inventory, the results are displayed beneath the container for the inventoried computer. The pub directory can be searched for matches on software packages and the resulting list of computers can be made into a computer group. You should run a fresh inventory on all clients before creating a computer group from inventory results.

To create a computer group from the inventory database:

  1. Click Search for Computers.

  2. Select an item to search for, such as a graphics adaptor or software package.

  3. Enter search criteria.

    You can specify the exact criterion to search for (for example, on a hard disk partition you can specify Bootable: yes, no, or either) or use the drop-down list to select the equal sign (=), less than or equal to sign (<=), or greater than or equal to sign (>=) to further qualify your search.

  4. Click Search. A list of systems matching the criteria appears.

  5. Enter a unique group name, then click Create Computer Group.

  6. The group is created in the container you were in when you performed the search. You can move the group to a different location if desired. (See Section 2.3.3).

When you create a group from inventory results, you do not need to link individual computers to this group. The links are created automatically.

4.4.2. Manually Creating a Computer Group

You can create a computer group based on any logical grouping you want. To manually create a computer group:

  1. Select the object container in which you would like to create the computer group.

  2. Select Create an Object.

    1. Click Object Type->computer group.

    2. Enter a name for the group and click Create.

    3. Click Create.

  3. Select the group name from the Contents pane.

  4. Select Add or Remove Links.

  5. Click Add.

  6. Enter a search string for the computer(s) you want to add.

  7. Choose the computers you want to link to this group by selecting the checkbox to the left of each computer name.

  8. Click Link.

You might construct a search string that can find all the computers you want to add at one time, or you might repeat steps 5-8 until all computers you want to add are located and linked.

4.5. Monitoring System Health

Caldera Volution Manager enables you to remotely monitor many client system resources. Resources include memory, disk space, inodes, processes, and other items that, if over-consumed, can cause your system to suffer performance degradation or crash.

You can implement one or more health policies to monitor these key resources. For each resource, you can define a threshold, and associate a special event with each threshold that causes an action to run. For example, you could:

  1. Configure a health monitor policy to measure free disk space on a partition.

  2. Specify in the policy that if free disk space falls below a certain level, a custom event named lowdiskspace is generated.

  3. Create an action named freespace to run any time the lowdiskspace event is generated. freespace could be a shell script that removes core files and other temporary files on the specified partition.

4.5.1. Creating a Health Policy

You can create as many health policies as you want. As long as the policies have unique names there is no limit to the number of policies you can define.

To create a health policy:

  1. Open the policies container.

  2. Select Create an Object.

    1. Click Create.

    1. Select Health Policy from the drop-down list.

    2. Enter a name for the policy (such as Monitor Desktops).

  3. Select the new policy from the Contents pane.

  4. Select Edit the Health Monitors.

  5. Click Add to add an entry.

  6. Select an item from the drop-down list, then click Add. Items are described in Section 4.5.1.1.

  7. Customize each entry. You need to enter:

    • any identifying information required, such as partition names or mount points

    • threshold values that cause notices (least severe), warnings (moderately severe), and alerts (most severe) to be generated

    • a name for any custom event you want generated when a particular threshold is crossed

    When done customizing the entry, click OK.

  8. Repeat step 7 for as many entries as you want to add.

  9. When you are finished configuring all entries, click OK.

  10. Link the policy to the desired computer(s) or computer group(s).

  11. Click Save Changes.

4.5.1.1. Editing Health Monitor Entries

The following resources can be monitored using a Health Policy. You can add a new health monitor or edit an existing monitor by selecting the Edit the Health Monitor task for a Health Policy.

Table 4-2. Health Monitor Entries

Health Policy EntryExplanation
CPU hog The percentage of CPU used by any single process. (not reliable for SCO OpenServer clients)
CPU utilization The percentage of CPU currently being utilized, defined as the sum of CPU percentage of each running process. Use loadavg for a more reliable measurement.
Free space on mount point The number of bytes that are free on a given mount point (for example, /usr/local).
Free space on partition The number of bytes that are free on a given partition (for example, /dev/hda2).
Free inodes on mount point The number of free inodes on a mount point (for example, /usr/local).
Free inodes on partition The number of free inodes on a partition (for example, /dev/hda2).
Free space in memory (bytes) The number of bytes of free memory. This is equal to free RAM + free swap space + buffers + cache.
Free space in memory (percentage) The percentage of total memory that is free. This is equal to freemembytes / (total RAM + total swap).
Load average The average number of jobs in the run queue, averaged over one minute. This is normally given in hundredths (a healthy load average is about 0.09). For Volution, they are multiplied by 100. A Volution load average of 9 is equal to a system load average of 0.09. A good health policy would be to set the threshold to 100* (number of processors). So, for a dual processor machine, a loadavg of greater than 200 would be worthy of attention. (not reliable for use with UnixWare or Open UNIX 8)
Memory hog (bytes) The number of bytes of memory that any single process uses.
Memory hog (percentage) The percentage of total memory used by any single process.
Number of processes The total number of processes in memory.
Number of named processes The total number of processes that match a given name.
Number of unique users The number of unique users currently logged on.
Number of users The total number of users logged on (not unique).
Is process in memory Is a given process currently running? (returns true or false)
Percentage of mount point used The percentage of used disk bytes on a given mount point (for example, /usr/local).
Percentage of partition used The percentage of used disk bytes on a given partition (for example, /dev/hda2).
Percentage of inodes used on mount point The percentage of used inodes on a given mount point (for example, /usr/local).
Percentage of inodes used on partition The percentage of used inodes on a given partition (for example, /dev/hda2).

4.5.2. Creating a Health Action

After you create a health policy, you must also create an associated action to run it.

  1. Open the actions container.

  2. Select Create an Object.

    1. Select Action from the drop-down list.

    2. Enter a name for the action (such as Desktop Health).

    3. Click Create.

  3. Select the new action from the Contents pane.

  4. Select Edit the Action.

  5. Select Check system health using linked Health Policies as the Action to run and click OK.

  6. Select Edit the Schedule and define a schedule (see Section 2.4.1.3).

  7. Link the action (see Section 2.4.2) to the same computers or containers to which you linked the health monitor policy .

  8. Click Save Changes.

4.5.3. Creating a Custom Action

Custom actions can be written to perform many different types of tasks. They take the same parameters as built-in actions, such as constraints, schedules, and links, but you provide the script to be run. You can customize an action to run automatically when Health Monitor thresholds are exceeded.

To create a custom action:

  1. Open the actions container.

  2. Select Create an Object.

    1. Select Action from the drop-down list.

    2. Enter a name for the custom action (such as freespace).

    3. Click Create.

  3. Select the new action from the Contents pane.

  4. Select Edit the Action.

    1. Select Run a script as the Action to run.

    2. Enter a valid script in the Script field. For example, a script to display a message to all users on each computer the action is linked to might be:

          #!/bin/sh
      wall << EOS
      The department NFS server is going down for maintenance in 5 minutes.
      Please save your work and log off now.
      Downtime will be 1 hour.
          EOS

      Any kind of script can be entered. The first line of the file indicates what shell or interperter is to be used (for example, #! /usr/bin perl, #! /usr/bin/python, #! /bin/csh, and so on). This has to be in the first line of the file, and the # character has to be the first character in the file. If no interperter or shell is given, the Volution agent assumes it is a shell script to be executed by /bin/sh, and it inserts a #! /bin/sh at the top of the script for you. For more information about shell scripting, see the operating system documentation.

    3. Click OK.

  5. Select Edit the Schedule and define a schedule (see Section 2.4.1.3).

    • If this action is independent of any custom event, select one of the time options.

    • If this action is meant to be run after a particular event, select Event and choose the event.

    • If this action is meant to be run after a custom event, select Event, select Custom Event, and enter the name of the custom event you created when configuring the health monitor policy.

  6. Link the action (see Section 2.4.2) to the desired computers or containers.

  7. Click Save Changes.

4.6. Managing Printers

Caldera Volution Manager enables you to manage printers on your Linux clients. You can modify your systems' /etc/printcap file to make printers available, set a default printer, or change printing options.

You can implement one or more printer policies depending on how networked and local printers are shared in your organization. If a group of clients access the same set of printers, you may want to create a policy for that group so that adding or removing printers can be easily managed.

4.6.1. Creating a Printer Policy

You can create as many printer policies as you want. For example, you might create one policy for each group of computers that share the same printers and default printer. As long as the policies have unique names and are linked to the correct computer or computer containers, there is no limit to the number of policies you can define.

To create the policy:

  1. Open the policies container.

  2. Select Create an Object.

    1. Select Printer Policy from the drop-down list.

    2. Enter a name for the policy (such as SalesPrinters).

    3. Click Create.

  3. Select the new profile from the Contents pane.

  4. Select Edit the Data.

  5. Enter information into the fields.

  6. When you are finished configuring all entries, click OK.

  7. Link the policy to the desired computer(s) or computer group(s).

  8. Click Save Changes.

4.6.1.1. Editing Printcap Entries

The following items can be configured using a Printer policy:

Table 4-3. Printcap Entries

Printcap Entry Explanation
Printer Device Printer device specification.
Remote Machine Remote machine's hostname.
Remote Printer Remote printer's name.
Spool Directory Spool directory pathname.
Max Job size Maximum job size in 1K blocks (0=unlimited).
Suppress Header Suppress the banner page.
Input Filter Default input filter program.
Baud Rate Serial port bit rate.
Description A comment identifying the printer for utilities like lpq.
Other AttributesAny other supported printcap attribute.

4.6.2. Creating a Printer Action

Once you create a Printer policy, you must also create an associated action to run it.

  1. Open the actions container.

  2. Select Create an Object. The Create Object tool appears.

    1. Select Action from the drop-down list.

    2. Enter a name for the printer action (such as UpdateSalesPrinters).

    3. Click Create.

  3. Select the new action from the Contents pane.

  4. Select Edit the Action.

  5. Select Update printcap file using linked Printer Policies as the Action to run and click OK.

  6. Select Edit the Schedule and define a schedule (see Section 2.4.1.3).

  7. Link the action (see Section 2.4.2) to the same computers or containers to which you linked the Printer policy.

  8. Click Save Changes.