System Analysis

System Analysis is intended as tool for detailed monitoring and analysis of a small group of systems. For this purpose, System Analysis comes with a set of predefined dashboards for common system types and offers the option to compose custom dashboards based on a large catalog of metrics available from different data sources. Flexible filtering and display options offer a variety of choices for graphical or tabular display. For detailed monitoring System Analysis offers to optionally collect fine-grained performance data like samples from ABAP work processes and HANA thread samples.

Available Data Sources

The following data sources can be consumed for displaying data:

Data Scource Description
System Monitoring (Monitoring & Alerting Infrastructure) All metrics available in System Monitoring (collected via the “Monitoring and Alerting Infrastructure”) can be consumed in System Analysis dashboards. Beyond this even metrics not visible in System Monitoring, but marked with the “RCA flag in the monitoring templates, can be used. Custom monitoring templates can be used in the same way.
Aggregated System Monitoring Data System Analysis provides a dedicated data store to retain a long-term history of selected metrics. The System Analysis dashboards seamlessly combine data from the System Monitoring store (which typically has a limited retention time to save space) and the aggregated store.
RCA / Hypercare There is a group of data stores for root cause analysis. Details are listed in a separate table below.
ABAP aggregated statistical records (“ST03 data”)

Collection of ST03 data (hourly and daily aggregates) is automatically scheduled as part of simple System Integration.

Note: This functionality is shared with “Real User Monitoring”

ABAP Exception Details

Details on various types of ABAP exceptions (short dumps, Gateway errors, Syslog errors) can be collected regularly and displayed in System Analysis.

Note: This functionality is shared with “Advanced Integration Management”

HANA Ad-Hoc SQL For HANA databases details can be retrieved ad-hoc via a pre-defined list of analysis statements.

Available RCA Stores

The following table shows which capabilities are available for intensive monitoring detailed Root Cause Analysis (hyper care):

CapabilityName Description SupportedSystem Type Collection
ABAP_WP_OVERVIEW ABAP work process overview /SDF/SMON ABAP Agent (Push)
HANA_THREAD_SAMPLES HANA thread samples HANADB Agent (Push)
HANA_MEMORY HANA component memory consumption HANADB Agent (Push)
HANA_LOAD_HOST HANA Loadgraph per host HANADB Agent (Push)
HANA_LOAD_SERVICE HANA Loadgraph per service HANADB Agent (Push)

HANA_EXPENSIVE_STATEMENTS

HANA expensive statements HANADB Agent (Push)
HANA_TABLE_SIZE HANA Table sizes HANADB Agent (Push)
SCP_NEO_OVERVIEW SAP Cloud Platform Neo application metrics EXT_SRV ESM (Pull)

Predefined Dashboards

For common system types pre-defined dashboards are delivered as part of the application. The subset of default dashboards actually shown in the list on the left side is filtered by the set of systems selected in the global scope selection. In addition, some dashboards are only shown and populated if corresponding data collectors are active for at least one system in scope. For example, HANA thread samples must be activated explicitly, and the page is only shown by default if the collection of HANA thread samples is active for one system in scope at the point in time when the global scope selection is done. 

The catalog of all available dashboards is also available in the “Page Catalog” node of the left side navigation. 

The following pre-defined default dashboards are available:

Name Content First Release Notes
ABAP on SAP HANA Combination of “ABAP Basis” and “SAP HANA 1.0 FP03 MAI-based
ABAP Basis SAP NetWeaver AS ABAP overview 1.0 MAI-based
ABAP Exceptions ABAP Exception details 1.0 FP02 EXM-based
ABAP Performance ST03-based data 1.0 FP02 ST03 data
ABAP Work Processes ABAP work process samples, drill down via guided RCA 1.0 FP03 Requires capability “ABAP_WP_OVERVIEW”
SAP HANA SAP HANA overview 1.0 MAI-based
SAP HANA Component Memory HANA memory details short-term and long-term 2.0 SP0 Requires capability “HANA_LOAD_SERVICE”
SAP HANA Tables HANA table statistics

2.0 SP0

Requires capability “HANA_TABLE_SIZE”
SAP HANA Thread Samples thread samples, drill down via guided RCA 1.0 FP03 Requires capabilities “HANA_THREAD_SAMPLES”, “HANA_EXPENSIVE_STATEMENTS”
AS Java SAP NetWeaver AS Java overview 1.0 MAI-based
Business Objects Business Objects Enterprise 1.0 MAI-based
Data Services Business Objects Data Services 1.0 MAI-based
Apache Tomcat   1.0 MAI-based

ASE

SAP ASE data base overview 1.0 MAI-based
SAP Cloud Platform Performance KPIs for “neo” 2.0 SP0 Requires capability “SCP_NEO_OVERVIEW”
Host Basic OS-level metrics 1.0  MAI-based

Release Notes

Focused Run 1.0 FP02

  • Exceptions view type

  • ABAP Performance view types (based on ST03 data)

  • Ad-hoc SQL statements for HANA

  • Metric forecast

  • Implicit selection of HANA and ASE databases associated to ABAP/Java systems

  • Improved metric selection dialog: suppress non-numerical metrics, improved performance, enhanced filtering options

  • Detailed tooltip when clicking on data point

  • Improved aggregation: View filter => expert

  • Unified configuration UI for aggregated data

Focused Run 1.0 FP03

Generic UI Enhancements
  • Auto-refresh via timer in header

  • Zoom to time range by dragging a rectangle on the chart (revert via wrench menu: Restore time range)

  • Page header bar for improved time navigation and scope selection. Also contains links to other applications and a status indication for System Anomaly

  • Links to other dashboards, other FRUN applications, and external tools 

  • Metric forecast: supportive pop-up to disable aggregation (e.g. instance->system), icon changed to align with System Monitoring

  • New configuration section “System Settings”: Allows to control data collection for “hyper care” for the systems in global scope selection, e.g. exceptions, ABAP work process samples, HANA thread samples

New RCA Stores
  • ABAP Work Process Samples

  • HANA Thread Samples

Predefined Content
  • New default pages:
    • ABAP Work Processes: supports drill down to task type and entry point 
    • HANA Thread Samples: supports drill down to thread state and application source
    • ABAP on HANA

Focused Run 2.0 SP00

Generic UI Enhancements
  • Tabular data display: Toggle data display between chart and table

  • "Technical Details" in Tools menu displays details on configured metrics and scope

  • Page Catalog listing all predefined pages

  • Cope with duplicate metrics, e.g. same technical metric name from different templates for same system 

  • Improved number formatting: reduce decimal digits for large numbers

  • Stable color assignment to systems / metrics cross views

  • UI5 cache buster activated to enforce automated browser cache refresh in case of updates

  • Improved time zone support: Aggregation to daily values now considers the browser time zone

  • Suppress click-tooltip for "Guided RCA" pages (ABAP WP Mon, HANA Thread Samples)

Personalization
  • Creatable view types are grouped by category (data source type): MAI, RCA, Statragg, etc.

  • View Personalization:
    • Separated into data source and display settings
    • Full scope selection for every view (select not only from global selection)
    • Top n filter for MAI
  • Personalizable links for views and pages (web, page, FRUN application)

  • Personalizable selection propagation (Guided RCA)

  • automated chart type adjustment for "no time axis" resolution

  • New chart option "show markers" to display data points. This helps to reveal "isolated" data points in charts with gaps

  • Metric catalog: new search function

  • Scaling of metrics

New RCA Stores
  • HANA Component Memory

  • HANA Table Sizes

  • HANA Statements

  • HANA Expensive Statements

  • HANA Load Host

  • HANA Load Services 

  • SAP Cloud Platform (Neo)

Predefined Content
  • New default pages:
    • SAP Cloud Platform
    • HANA Component Memory
    • HANA Tables
    • Hosts page: top 5 file systems per host instead of virtual CPU
    • ASE page: enhanced to fill all charts, no matter if ABAP-based or BOBJ-based
    • HANA Thread Samples: tooltip of the bottom graph displays the HANA sql statement in full text

 

Prerequisites

General

  • System Analysis is available as of Focused Run 1.00. Up to Focused Run 1.0 FP02 the application was called “System Analytics” and part of the “Advanced System Management” use case.

  • System Monitoring Setup of the Technical Systems involved is required for the default dashboards to be populated

Technical Requirements - Relevant SAP Notes

Focused Run 1.0 FP02
  • 2608066 - System Analytics: Corrections for FRUN FP02

Focused Run 1.0 FP03
  • 2653422 - System Analysis: Corrections for FRUN FP03

  • 2639985 - Collective corrections for FRUN UI5 applications 1.0 FP03

  • 2555664 - Enable HANA Dashboards and AdHoc SQLs in System Analytics for Focused Run

Focused Run 2.0 SP0
  • 2555664 - Enable HANA Dashboards and AdHoc SQLs in System Analytics for Focused Run

Setup

For basic usage of System Analysis no dedicated setup steps are necessary. All systems for which System Monitoring is configured should automatically also work in System Analysis.

System Analysis Aggregated Store

The System Analysis aggregated store allows to retain selected metric data from the MAI store beyond the standard retention time of the MAI store. The content of the aggregate store is defined based on variants. Using these variants a daily task collects and aggregates data from the MAI store and persists it in a dedicated store.

For each variant you can define

  • The retention period (in days): After this period data will be eliminated from the aggregate store

  • The lookback time (in days): This is the overlap of the MAI store with the aggregate store. For a MAI store retention time of 28 days and a lookback time of 2 days the aggregation task will gather data that is 26 days old.

  • The aggregation granularity: You can choose between 5 minutes, 15 minutes, hourly, and daily granularity. Choose you appropriate compromise between granularity and space consumption.

  • Customer networks: You can restrict variants to selected customer networks

  • Data store: “Standard Aggregate Table Store” is the only option supported for consumption in the System Analysis application.

  • Systems: You can specify various filter criteria to select for which systems data aggregation should take place at all

  • Metrics: You can select metric categories and individual metric names for aggregation

Note that there is no mechanism to automatically detect and warn about overlap between variants. Technically it is possible to capture the same metrics for the same systems in different variants. Make sure to carefully tailor the variants for your needs. The “Calculate Duplicates” button allows to estimate overlap of a variant with all others. You can also decide to intentionally collect identical metrics in different variants, e.g. with a daily granularity for a long retention period and in addition with a finer granularity with a stricter retention policy.

ABAP Performance Data

Periodic collection of ABAP performance data is automatically scheduled as part of SSI. As pre-requisite the aggregate performance data must be available e.g. in transaction ST03N in the managed system. A typical source for missing data is that the corresponding aggregation job is not active in the managed system. See Note 2369736 for troubleshooting missing data in ST03N. The “related links” in the System Analysis application points to the “ABAP Performance KPI Setup” application. Here you can explicitly select the instance to use for data collection and check the collection status.

Root Cause Analysis Data Collection

ABAP_WP_OVERVIEW

You can activate /SDF/SMON in the managed system from within System Analysis. If an analysis is already scheduled in the managed system by another user the existing scheduling is not changed. 

Following details can be scheduled for analysis in managed system additionally:

  • SM04 = Memory per (user) modes

  • ST06 = CPU and paging activity

  • TOPCPU = Top CPU processes

These non-default flags have a higher impact on the managed system and the data is not collected and transferred to FRUN but only helpful for local analysis in managed system. 

In the managed system the batch jobs /SDF/SMON_SCHEDULER and /SDF/SMON_WATCHDOG are scheduled which are responsible for the local measurement.

HANA SQL scripts

To use the RCA capabilities for system type HANADB you need to copy the scripts from note 2555664 (Enable HANA Dashboards and AdHoc SQLs in System Analytics for Focused Run) to all host agents of your HANA databases.

HANA Expensive Statements

To collect the expensive statements from the HANA DB they need to be enabled in the HANA database first. See also https://help.sap.com/viewer/bed8c14f9f024763b0777aa72b5436f6/2.0.03/en-US/5faf04f17830464eacdb7938b383d2ab.html or FAQ note 2180165 (FAQ: SAP HANA Expensive Statements Trace) 

SAP Cloud Platform Metrics

To enable the collector for SAP Cloud Platform (Neo) metrics an external service with the required http endpoint and root URL needs to be created. 

Open the Cloud Service Management Configuration and add an external service of type SAP Cloud Platform (Neo) for your account with the root URL https://api.hana.ondemand.com. Create an endpoint with your user and a proxy if required.

 

Using the Application

When launching System Analysis either from the FRUN Launch Pad or via an explicit link you will first receive a popup to choose a set of systems to look at. By default, the number of systems to be displayed in System Analysis is limited to a maximum of ten (see FAQ section to raise the limit and to avoid the initial scope selection). If you choose more than ten systems in the bottom table of the scope selection, then the first ten will be chosen.

After finishing the scope selection, the System Analysis application comes up with the typical layout common to most Focused Run applications:

The header bar contains a set of global tools:

  • “Hamburger” menu to expand the page names on the left side
  • Back button for page navigation
  • Refresh timer for automated update of the current display
  • Help icon for context-sensitive help (if configured)
  • Calendar icon for global time selection
  • Target icon for global scope selection
  • Configuration icon to expand the configuration section. In System Analysis you can configure data collection per system in scope and RCA capability and custom web links.
  • Message icon for notifications
  • Personalization icon

The list of pages (dashboards) on the left side will be populated with default pages (based on selected system types), the page catalog, and custom pages.

The page toolbar serves as overview and selection tool for the content of the page. It consists of the following main sections:

  • Time Navigation: You can explicitly set the time range for the current page, scroll via the arrow buttons, or choose one of the pre-defined relative time ranges.
  • Scope refinement: For default pages the subset of systems for which the page is applicable is automatically assigned to the page and displayed in the toolbar. To focus on a single system, refine the scope here without touching the global selection.
  • Links: A set of links to other pages and other Focused Run applications may be available in the “links” button. For custom pages you can choose the links. Where possible the current scope selection of the page is propagated to the link.
  • System Anomaly: If System Anomaly Analysis is configured and available for the systems in scope then the number of systems for which an anomaly is expected is shown here. Click on the button to show details or navigate to System Monitoring for even more details.

Single View Operation

Every single view provides the following functionality. Depending on the capabilities and data sources consumed not all aspects will be available for all views.

General Chart Functionality
  • Hover over a single data point to get a short tool tip that will be closed automatically.
  • Select / click on a single data point to get a more detailed pop-up view on the single data point and the series. The pop-up may contain links to related tools for deeper targeted analysis. Where appropriate the context of the selected data point (e.g. one out of several hosts) is propagated.
  • Drag a rectangle over more than one data point to zoom into a time range. Resolution is adjusted automatically (if set to “auto”). This procedure can be repeated multiple times, down to the granularity provided by the backing data store. The zoomed time range can be reset to the page-level selection or propagated to all charts on the page via tools menu items.

View Title Bar

The following options are available via the View title bar:

  • Forecast (for MAI-based data only): Based on the historical data the future development of a metric is predicted and indicated as a dashed line. If the shown metric is based on aggregation (e.g. aggregating from all instances of a system to a single metric) then you will first be prompted to disable aggregation. The forecast can be removed by toggling the button again.
  • Table/chart toggle: Data display can be switched between tabular and graphical display.
  • Filter: The filter menu allows to transiently override the selection of displayed data. All changes here are not persisted and gone upon restart of the application:
    • Time range: De-couple from page selection
    • Resolution: override the granularity
    • Aggregation: Review which dimensions should be aggregated. Aggregated dimensions disappear effectively from the chart.
    • Metrics selection:
  • Export: Data can be downloaded in CSV format for offline analysis
  • Links: view-specific links
  • Technical details: provides details like MAI GUIDs for detailed analysis related to MAI templates etc.

Guided Root Cause Analysis

There are currently two pre-defined dashboards that provide built-in drilldown into performance data: ABAP Work Processes and HANA Thread Samples. The idea is that selecting a data point in a chart (by clicking on it) serves as filter criteria for the subsequent chart. As a concrete example for ABAP Work Processes the following might be a useful workflow:

  1. In the first chart, zoom into a time range by dragging a rectangle -> the chart will update to the new time range
  2. In the first chart click on a single data point: The second chart will be published now using the selected work process type and time stamp of the data point from the first chart as filters. The time stamp is indicated via a black dotted vertical line. Around this time stamp 10 data points are rendered (if available). As additional dimension now, the entry points for the select task types (e.g. program name) are shown.
  3. In the second chart click again on a single data point. This triggers population of the third chart with the filters time stamp, task type, and entry point. The third chart shows the actions for the selected entry point.
  4. You can adjust the aggregation of dimensions of the third chart via the filter button. For example, switching off the aggregation for “User Name” and instead activating aggregation for “Action” may reveal the actual user names (assuming the proper authorizations are assigned)

The Guided RCA approach can also be applied to custom dashboards, e.g. by copying a predefined dashboard and adjusting the “Selection Propagation” settings in the view personalization.

Custom Dashboards

You can create custom dashboards by copying existing dashboards or by starting via the “Add Page” functions in the left-side navigation or in the personalization. The mechanisms are basically like all other Focused Run applications, thus only application specific details are explained here.

Procedure to Create a System Analysis Dashboard:

  • Create a page, assign name, icon; decide if you want to make it “public”
  • Decide on the layout of the page, e.g. grid layout
  • Drag views from the available view types onto your dashboard. Via the view type you basically decide what data source should be used, .e.g, MAI data, RCA data, ST03 data, etc.
  • Personalize each view by clicking on the “personalization” icon on the view title bar. Depending on the view type the flexibility on personalization options may vary.
  • Save the personalization settings to persist the data. The new dashboard will reappear every time you launch the application.

Personalize View – Data Source Settings:

  • Choose the set of systems if you want to override / hardwire the scope selection. Otherwise the global scope is applied.
  • Choose the capability to consume (for RCA) and the metrics from the metric catalog.
  • In the metrics dialog, tab “Selected”: assign metrics to a second axis if needed. Apply scaling if needed.
  • In the metrics dialog, tab “Aggregation”: Decide which dimensions should be aggregated by default, set filters if needed.
  • If desired override the time range.
  • Resolution: You can decide to switch off the time axis completely (“No Time Axis”) or to explicitly set a resolution. Note that actually used resolution may also depend on the granularity of data collection. When switching off the time axis, time-series charts cannot be used (see chart settings).
  • Top N / Top Rank: The RCA data source provides filtering. 0 means no filter.

Personalize View – Rendering Settings:

  • Choose the initial rendering type: chart or table
  • Set the view title
  • Decide if you want a legend displayed
  • Data labels are rendered as texts displaying the values of data points in the chart
  • Markers are rendered as “dots” for every data point. If a chart contains too many data points the markers may still not be visible. Use markers e.g. to expose isolated data points that are surrounded by “gaps”.
  • Chart type: time series charts are only available if the resolution is not set to “No Time Axis”. On the other hand side, pie and donut chart types only make sense if you set the resolution to “No Time Axis”

Personalize View – Link Settings:

  • Choose the set of links that you want to have displayed in the view tools menu and in tooltips. For Web links you might want to maintain a global list of links from which you can choose in Configuration -> Web Links.
  • In Configuration -> Web Links you can maintain any web links that make sense in your organization. Any link maintained here is visible to all users.
  • For web links you can use the following place holders that will be filled by the current context (system/host) if possible:

Placeholder Description
{SYSTEM.ExtendedSID} Extended sid, e.g. FRNADM
{SYSTEM.SID} Short 3-letter System ID, e.g. FRN
{SYSTEM.SystemType} System type like “ABAP”, “JAVA”
{SYSTEM.NamespaceName}  
{SYSTEM.LMDBID} LMDB guid (32 char notation)
{HOST.LMDBID} LMDB guid (32 char notation)
{HOST.NAME} Unqualified host name