Jump to content English
hp.com home
HP Global Workload Manager (gWLM)  |  View Real-time Reports

View Real-time Reports

  HP Global Workload Manager (gWLM)

»gWLM concepts / overview
»Getting the most out of gWLM
»Getting Started - gWLM Home
»Manage New Systems (Create SRDs)
»Manage Workloads
»Unmanage Workloads
»Edit SRDs
»Edit Policies
»Edit Workloads
»Edit Associations
»Configure Events
»View Summaries
View Real-time Reports
»View Historical Reports
»Set Up Read-only Monitoring
»Web browser tips
»Tab: Shared Resource Domains
»Tab: Policies
»Tab: Workloads
»Tab: Associations
»Glossary
Content starts here

Reach this window by selecting:

Optimize Global Workload Manager (gWLM) View Real-time Reports

gWLM allows you to monitor workloads and policies in real-time by selecting the View Real-time Reports menu item. Each refresh of a real-time graph includes values updated from when gWLM last changed CPU allocations.

From the View Real-time Reports window, you can:

Frequent changes in allocation indicate that gWLM is optimizing the amount of CPU resources available to a workload's compartment, benefitting the workload.

The interval at which graphs are refreshed is not related to the interval on which gWLM changes CPU allocations.

The View Real-time Reports window provides a table that includes the following columns:

ColumnDescription
WorkloadName of a workload
PolicyPolicy associated with the workload
Policy Reading

For utilization policies and OwnBorrow policies: The reading is utilization.

For fixed policies: The reading is the amount of CPU resources guaranteed to the compartment of any associated workload.

For custom policies: The reading is the value of the metric you update using the gwlmsend command.

Compartment Utilization

The compartment consumption divided by the compartment's size

For instance, if a compartment's consumption is 2 CPUs and its size is 4 CPUs, the compartment utilization is 50%.

ModeManaged / Advisory
HostnameHost where workload is running
Compartment TypeType of compartment in which workload is running
Shared Resource DomainSRD in which compartment is located

Using the table:

  • Re-sort the rows by selecting the column heading that you would like to sort by.

  • Limit the items displayed by using the Filter by feature above the table.

Modes

You will see one of the following mode indicators directly above the graphs:

  • ( Advisory Mode )

  • ( Managed Mode )

These indicators specify whether the data being graphed reflects an SRD that was deployed in advisory mode or managed mode.

Buttons: |<, <, >, and >|

The following buttons are available for panning graphs:

ButtonDescription
|<

Moves to the beginning of the data

This button is grayed out if you are already at the beginning of the data or if the amount of data does not yet exceed the graph range.

<

Moves back one graph range

This button is grayed out if the amount of data does not yet exceed the graph range, or if you are already at the beginning of the data.

>

Moves forward one graph range

This button is grayed out if the amount of data does not yet exceed the graph range, or if you are already at the end of the data.

>|

Moves to the end of the data

This button is grayed out if you are already at the end of the data or if the amount of data does not yet exceed the graph range.

You can adjust the following items as indicated:

Graph range

The graph range determines how much data the graph shows.

The default range is 15 minutes.

You can change this range by editing the following property in the file /etc/opt/gwlm/gwlmcms.properties on your gWLM CMS:

com.hp.gwlm.ui.monitor.viewport

Restart gwlmcmsd for a change to this property to take effect.

Real-time cache

This cache size determines the number of minutes of data to maintain on the managed nodes for the purpose of generating real-time reports. (To minimize network traffic, the managed nodes send the full day's data only once per day.)

The default number of minutes is 20.

You can change this range by editing the following property in the file /etc/opt/gwlm/gwlmcms.properties on your gWLM CMS:

com.hp.gwlm.node.cachesize

Deploy or re-deploy a given SRD for a change to this property to take effect for that SRD.

View Workload Detail

Select a workload in the table then the [View Workload Detail] button to see:

Workload parameters are described in the table below.

ParameterDescription
HostnameHost where compartment is located
Compartment TypeThe type of compartment: npar, vpar, pset, or fss group
ModeManaged / Advisory
Compartment Maximum CPUsThe upper bound on the number of CPUs the compartment can have, based on the compartment type, the number of compartments, and the number of CPUs in the complex
Initial Compartment CPUsThe number of CPUs the compartment had when gWLM discovered it
Compartment Minimum CPUsThe lower bound on the number of CPUs the compartment can have

Select the [Back to Workload Selection] button to return to the window listing all workloads. You can then select another workload to graph, if desired.

Workload Utilization graph

This graph shows a workload's allocation and its utilization of that allocation.

The graph labels each axis as follows:

  • Left vertical axis: # CPUs

  • Horizontal axis: Time

  • Right vertical axis: % Utilization

Labeling the two vertical axes differently allows for the presentation of additional data.

You can graph each of the following items by selecting its checkbox below the graph:

Graphable itemDescription
Allocation

The amount of a resource, such as CPU, that gWLM sets aside for a compartment after arbitrating resource requests from the policies for all the compartments.

In managed mode, gWLM makes an allocation available to a compartment. In advisory mode, however, gWLM simply reports what the allocation would be--without actually affecting resource allocations on a system.

Size

The amount of a resource a compartment actually has.

For CPU, size may differ from the allocation when gWLM is deployed in advisory mode.

CompartmentMin

The minimum amount of a resource that a compartment can have. As an example with the CPU resource, CompartmentMin is the number of CPUs a compartment can have based on how it was defined when created using npar, vpar, or pset commands.

gWLM's discovery feature evaluates a system to determine this value.

CompartmentMax

The maximum amount of a resource that a compartment can have. For example, with the CPU resource, CompartmentMax is the number of CPUs a compartment can have based on how it was defined when created using npar, vpar, or pset commands.

gWLM's discovery feature evaluates a system to determine this value.

Consumption

The amount of a resource being consumed by the processes in a compartment.

For instance, with the CPU resource, compartment consumption is the amount of CPU being used by the compartment's processes.

Utilization (right axis)

The percentage resulting from dividing a workload's consumption by its allocation.

For example, assume a workload has an allocation of 5 CPUs. If the workload is consuming 4.5 CPUs, its utilization percentage is 4.5/5, or 90%.

View Policy Detail

Select a policy in the table then the [View Policy Detail] button to see:

The policy parameters shown above, plus the parameters for fixed policies and OwnBorrow policies, are described in the table below.

ParameterDescription
Policy TypeType of policy: Fixed, Utilization, OwnBorrow, or Custom
Policy PriorityThe priority of the policy, with 1 being the highest priority (This parameter does not appear for fixed policies or OwnBorrow policies, as they do not have priorities.)
TargetThe target value for the metric (for custom policies)
Target UtilizationThe target utilization percentage (for utilization and OwnBorrow policies)
Target SizeThe target CPU allocation (for fixed policies)
Policy Maximum CPUsThe largest amount of CPU resources a policy will request for its associated compartments, as specified in the policy definition
Policy Minimum CPUsThe smallest amount of CPU resources a policy will request for its associated compartments, as specified in the policy definition
Policy Owned CPUsThe amount of CPU resources a policy attempts to guarantee for its associated compartments when needed, as specified in the policy definition

Select the [Back to Policy Selection] button to return to the window listing all workloads and policies. You can then select another policy to graph, if desired.

There are two policy graphs:

Policy Request and Workload Allocation graph

This graph shows what allocation the policy is requesting for the workload's compartment and the allocation the compartment actually gets after gWLM arbitrates the requests from all the policies.

The graph labels each axis as follows:

  • Left vertical axis: # CPUs

  • Horizontal axis: Time

You can graph each of the following items by selecting its checkbox below the graph:

Graphable itemDescription
Request Size

The amount of a system resource that a policy asks gWLM to give to the policy's compartment. (Parameters you specify in defining a policy may restrict its request.)

Allocation

The amount of a resource, such as CPU, that gWLM sets aside for a compartment after arbitrating resource requests from the policies for all the compartments.

In managed mode, gWLM makes an allocation available to a compartment. In advisory mode, however, gWLM simply reports what the allocation would be--without actually affecting resource allocations on a system.

The amount allocated may be restricted by the minimum and maximum number of CPUs you specify for compartments when creating them using npar, vpar, or pset commands.

Size

The amount of a resource a compartment actually has.

For CPU, size may differ from the allocation when gWLM is deployed in advisory mode.

Owned

The amount of owned CPU resources. This item appears only when graphing OwnBorrow policies.

Effective PolicyMin

PolicyMin is the minimum amount of a resource, such as number of CPUs, that a policy will request for any compartment with which it is associated. Effective PolicyMin is the larger of PolicyMin and CompartmentMin (the minimum amount of a resource that a compartment can have.)

Effective PolicyMax

PolicyMax is the maximum amount of a resource, such as number of CPUs, that a policy will request for any compartment with which it is associated. Effective PolicyMax is the smaller of PolicyMax and CompartmentMax (the maximum amount of a resource that a compartment can have).

Policy Input and Resulting Request graph

This graph shows the input to a policy (a "measured" value or a "reading" value and a "target" value) and the resource request that is generated based on that input.

The graph labels each axis as follows:

  • Left vertical axis: % Utilization

  • Horizontal axis: Time

  • Right vertical axis: # CPUs

Labeling the two vertical axes differently allows for the presentation of additional data.

You can graph each of the following items by selecting its checkbox below the graph:

Graphable itemDescription
Measured Utilization*

The current value of a metric being used in a policy. gWLM calculates utilization metrics for you.

Measured Size**

The size of a compartment in CPUs.

Reading***

The value of the metric being used in a custom policy.

Target Utilization*

The value that drives a policy, influencing its resource requests to gWLM.

For a target CPU utilization, gWLM attempts to keep a workload's CPU utilization below the target by adding CPU resources when the workload is using too much of its current CPU allocation. For example, assume a workload has a utilization policy with a target of 80% and a size of 5 CPUs. If the workload is consuming 4.5 CPUs, its utilization percentage is 4.5/5, or 90%. gWLM would attempt to allocate additional CPU resources to the workload to meet the target. A size of 6 CPUs would result in a utilization percentage of 4.5/6, or 75%, meeting the target.

A target can also be a value the workload should stay above, such as x transactions per second. Adding resources in this case helps the workload maintain the number of transactions.

Target Size**

The desired size, in CPUs, of the workload. This value drives a policy, influencing its resource requests to gWLM.

Target***

The value the metric (graphed as 'Reading') is being compared against. This value drives a policy, influencing its resource requests to gWLM.

Request Size

The amount of a system resource that a policy asks gWLM to give to the policy's compartment. (Parameters you specify in defining a policy may restrict its request.)

* Utilization values are graphed for utilization policies and OwnBorrow policies.

** Size values are graphed for fixed policies.

*** Reading and Target values are graphed for custom policies.