Capacity Planner Release Notes 99


Overview

Released: December 2018

This release introduces SSO support using SAML 2.0 allowing for seamless integration with your identity provider. There has also been a consolidation of generic reporting. This removes the amount of overlap between reports, provides consistent input parameters where appropriate and allows for the same report to be run across different ‘entity types’ (for example, physical servers, virtual machines, hosts, etc.). Finally, improvements have been made to benchmark processing to ensure that consistency is provided across common chip sets.

Highlights

Single Sign On (SSO)

Single Sign On allows Capacity Planner user authentication to be validated within your organisation using SAML 2.0. When used, it will allow you to log in to Capacity Planner using your existing SSO credentials that are used for other applications across your organisation. There is no need to register with Capacity Planner to make use of this and, upon successful authentication, you will have access to your organisation’s Capacity Planner project automatically.

Making use of SSO and configuring the system to operate correctly with your Identity Provider requires some specific configuration and completion of a small end to end test for verification purposes. Should you wish to make use of this feature, please contact your account representative.

Reporting features

Prior to this release, there were a number of reports, many of which performed the same reporting activity, but simply on a different type of device or managed entity. Also, the configuration of input parameters for some reports were not as flexible as others. For example, some reports may allow you to run over a chose time frame as identified by a user defined baseline, while others would not.

Several changes have been made to the standard set of reports to firstly reduce the amount of reports available but make those reports more flexible and powerful.

Filtering

There are several standard parameters now available for most of the reports where appropriate. Most reports allow the user to select a number of entities. An entity is a component managed by capacity planner and, by default, can be either a Cluster, Host, VM, Physical Server or a Datastore. When the user selects an entity, they will have the facility to filter that selection of entities using a grouping filter, Grouping Filter [1]. In this drop down, the user can specify whether to select all entities of that type, choose from a list or select a grouping value. If the user chooses to select a Grouping Filter, then use ‘Grouping Value [1]’ to select the filter. The amount of entities that have a particular grouping value will be shown in the list. If the user chooses to select from a list of names, use the ‘Level 1 Entity Name’ drop down.

After doing this, the user can further filter. If the user wishes to add another grouping filter, but remain using the same entity (for example, select all VMs that are Powered On and part of the ATM service), simply choose the same Level 2 Entity Type and repeat the above process. However, the user can also select entities that are contained within the chosen Level 1 entity. For example, if Level 1 entity was a named Cluster, then if Level 2 Entity was chosen as ‘VM’ then the report would be filtered to all VMs within the chosen Cluster. If the Level 1 Entity type was chosen as ‘VM’, then the user could use the Level 2 Entity Type filter to choose drives of those VMs. Again, this can be further filtered using groupings or selections from a list.

The filtering mechanism allows for 3 levels of filtering which should be sufficient for reporting in Capacity Planner.

Metrics

The ‘Metric’ Filter allows the user to select which metric to report on. Some reports allow for multiple metrics. Only those metrics available on the lowest level entity in the filtering selection will be available.

RAG Thresholds

The Red and Amber Operators and Threshold options allow the user to select thresholds for the chosen metric(s).

Summary Level

Summary Level allows the user to select the range of measures used for the metric. Baseline summary provides a set of measures covering the entire time period. This is the fastest type of report. Monthly summary provides a set of statistical measures per calendar month for which there is data. Daily summary provides a set of measures per day. Unaggregated data will run a comparison against the entire time period. Careful consideration should be used before running an unaggregated report as this can take a considerable amount of time to complete if a long date range is used across multiple entities.

Date Range Type

This can either be a relative date range or a specific date range. Relative date ranges allow the user to choose a range for the report relative to the current day. When using this, the Relative Range field provides the facility to use from latest complete months, weeks or days. The Date/Time Periods then specifies how many of these complete date ranges to use. For example, the user can select a relative date range of ‘Latest ‘N’ Complete Months’ then entering ‘2’ in the Date/Time Periods field, define how many complete months.

When using complete months or weeks, only complete calendar months or weeks will be considered. For example, if the report is run on the 7th of December for the last relative complete month, only data for the month of November will be included.

Specific range allows the user to enter specific dates in the Date from and Date To fields. Please note that the date must be entered in US format (Month/Day/Year).

Chart Type

Select whether to use a box plot, line chart or both. For the box plot, all measures for the chosen summary level will be used. For a line chart, only the chosen percentile measure will be plotted on the chart.

Percentile/Value

The measure of the metric(s) to use in the report for display and threshold comparisons.

Y-Axis Scale

Choose whether to fix this scale from 0-100 or show all values or scale to fit. This will often update automatically depending on the metric.

Legend

Per report flexibility provided on the way the legend is presented on the screen

Report Size

Some reports may take a long time based on the complexity, number of data points in the query and comparisons to be made. This allows the user to specify whether they are willing to wait on the report to run to completion. If a report data set is larger than the selected value, the user will be alerted and have the option to increase the allowable report set size.

The following is a screenshot showing an example filter. In this example, the user has created a report for CPU Ready % per VCPU metric. The user wants to know if the CPU Ready % per VCPU metric for any SQL VM in the Consumer Business Area has had a single day in the last calendar month where it has spent 5% of its time above a value of 5 for amber or 10 for red.

Standard reports

The standard reports available are described below.

Cluster events

This report remains unchanged. It will show all current projected cluster capacity events and the time series for that event. For each event, the user can click to view history of that event. A chart shows how often that event has been raised and whether that event is getting closer over time or moving further into the future. A time series of allocated vCPUs in that cluster is also provided.

In the screenshot we see an example cluster event showing increasing CPU over time. Should this continue, the cluster will soon have less than one redundant server available.

Cluster Events History

With this report user can select a cluster and metric combination and see the history of any projected cluster events. This can also be accessed directly from the Cluster Events report.

In the screenshot we can see an example of an output of this report. Showing CPU and Memory daily box-plots with the RED warning set to 80% and AMBER warning set to 70%. Report is ordered by RAG status.

Metric Analysis Multi-Chart

Allows the user to create a RAG chart identifying when a combination of metrics for a selection of entities has breached a threshold condition. A separate unique chart is drawn for each entity/metric combination. The chart is order by RAG status.

In the screenshot we can see an example of single chart reporting, showing a time series of hourly average CPU for the previous full calendar month.

Metric Threshold Analysis

Provides the facility for the user to filter on entities and select up to 3 metrics for threshold breach checks. This is presented in a visual tabular format of horizontal bar charts. A bar will be shown as ‘red’ if a threshold has been crossed. Clicking on the bar will allow the user to investigate the timeseries data for that entity/metric combination.

In the screenshot we can see an example report output of peak CPU, Memory and CPU Ready% per VCPU over an entire baseline period.

Metric Thresholds Percentiles

Allows the user to create a table of all the percentile values for a metric where it has breached a given threshold (which can be set to 0 to view all entities). This is particularly useful for identifying very busy or inactive servers. Clicking on an entity name allows the user to view box plots and time series data for that entity.

In the screenshot we can see an example output showing all VMs where the 95th%ile of CPU is greater than a specified value. In this case, using 0 will show all measures for this metric for all machines.

Scatter Plot

Draws a scatter plot when provided with two entity and metric combinations.

Time Above Threshold

Will highlight when a metric for a chosen entity time series data has spent any time above a threshold. It will determine the total time spent above that threshold and the longest period of time. The user can click through to visibility into the time series and the longest threshold breach.

VM Storage Summary

Provides a detailed report showing the comparison between storage allocated to a virtual machine and storage used by the virtual machine OS drives. Can be used to identify potential storage waste in an environment. Is ordered by most wasteful VMs.

Benchmarking Improvements

A more automated process has been put in place to apply benchmarking scores to servers as they are discovered. The previous report could result in slight inconsistencies between benchmarking scores for machines with the same chipset. This will ensure consistency across all benchmarking.

Fixed issues

There was a bug where if there VMs which had no value for CPU or Memory were contributing a demand value of 0 for these metrics to the average VM, when they should have not been contributing at all. This has been resolved.

When changing groupings quickly, it was possible for the sunburst to disappear from the UI. Closing the browser tab and then re-opening would resolve this. However, the condition that caused this issue has now been resolved.

Other releases

Click the links below to view other versions of Capacity Planner release notes.

Capacity Planner Release Notes 101 Released: August 2019
Capacity Planner Release Notes 100 Released: May 2019
Capacity Planner Release Notes 99.2 Released: March 2019
Capacity Planner Release Notes 99 Released: December 2018
Capacity Planner Release Notes 95 Released: May 2018
Capacity Planner Release Notes 94 Released: March 2018
Capacity Planner Release Notes 93 Released: January 2018
Capacity Planner Release Notes 92 Released: December 2017
Capacity Planner Release Notes 91 Released: October 2017
Capacity Planner Release Notes 90 Released: September 2017