Real Time Monitoring and Historical Reporting
In Eagle's Automation Center, you can create and maintain technical and business SLAs in the SLA Management area. In the Automation Desk area, you can monitor SLAs for scheduled jobs and events that ran within a 24-hour period, midnight to midnight, of any given day.
In System Management Center, you can run daily, weekly, and monthly SLA historical reports via system queries located in the Diagnostics area. The SLA historical data is stored in a separate database table and can be easily extracted any time for comparison.
Actual Performance Indicators
Service Level Agreements (SLAs) use Key Performance Indicators (KPIs) to help you measure and improve operational efficiency. KPIs are defined and assigned to processes based on the process type. Upon completion of a scheduled process, KPI execution statistics are captured so you can monitor the process SLA and generate SLA status reports for daily, weekly, or monthly performance.
The following KPIs are supported for selected processes: Duration, Start/Stop Time, Records per Second, Total Record Count, Processed Record Count, and Failed Record Count. The following KPI aggregates are also supported: Minimum, Maximum, Sum, and Average.
System Metrics-Based Framework
The SLA Management framework allows you to define, capture, calculate, and report on various technical and business system performance measurements. You can capture information from various data sources, define and set targets, measure performance, and monitor activity as well as produce historical reporting. This framework uses a thin metadata layer wrapped around the existing database and application logs, leveraging this extensive set of system metrics data.
With the SLA Management framework, you can:
- Capture. System metrics are generated and captured by the dedicated execution logs of applications and processes, for example, execution logs of scheduled jobs or message streams.
- Define. KPI attributes, KPIs, and SLA thresholds are combined to create benchmarks for technical or business KPIs.
- Calculate. Captured system metrics are evaluated against the benchmarks for a specified time period. Typically, this is a simple comparison, but may be more complex.
- Report. Two types of reports are available: current day reporting (monitoring) in Automation Center's Monitor workspace and historical reporting via the System Management Center queries that can evaluate captured metrics against SLA thresholds across a specified period of time.
SLA Framework Terminology
Data Source
A data source is a set of system and user-defined database metric views layered around the captured Eagle data. Eagle supports SLA management for the following types of metrics: Business, Reference Data, Process and Job, User, Operational Performance, Database Engine, and Platform.
Source Type
A source type is a loosely-defined data source where system and job statistics are stored. This loose coupling enables an extensible and open-ended set of sources that can be used by SLA Management. Some of these sources are included out-of-the-box in each release, (for example, 2015 R2 contains the Schedules, Message File, and Message Stream sources) while others are user-defined (via a simple SQL view) for further customization. Examples of out-of-the-box-and user-defined source types include: Schedules, Message Streams, Message File, Vendor Data, Activity Logs, Database Logs, etc.
Measure Type (data format)
A measure type is a specific metric that can be extracted from the source type. Currently supported measure types include: Duration, Start Time, Stop Time, Throughput (Records per Second), Total Record Count, Processed Record Count, and Failed Record Count.
KPI
A KPI is an instance of a given KPI attribute, referencing the metrics from a specific data source. For example, using the Uploader Duration KPI attribute for a specific source name, FX Price Load, you can create a new KPI - FX Price Load Duration - that can be used in SLA calculations. KPIs measure performance against the SLA benchmarks as time passes.
The following KPI aggregates are currently supported: Minimum, Maximum, Sum, and Average.
KPI Attribute
A KPI attribute defines the basic KPI building blocks, such source type and measure type. An example of a KPI attribute is an Uploader Duration, where the source type is Uploader and the measure type is Duration. A KPI Attribute is made up of Source Type and Measure Type and a KPI is made up of a KPI Attribute and a Source Name.
SLA
An SLA is a definition of threshold settings associated with one or more KPIs. SLA sets benchmarks for current KPIs and is key in measuring future performance. SLAs are often confused with KPIs, but it's important to remember that an SLA is designed to measure future performance whereas a KPI focuses on past performance.
| Info | ||
|---|---|---|
| ||
| A KPI may be used in several SLAs. An SLA may reference several KPIs in which case aggregate calculations will be used. |
SLA Group
An SLA group allows you to easily manage and migrate related business or technical SLAs.
SLA Data Relationships
The following are supported SLA data relationships:
- KPI-Attribute-to-KPI. KPI is made up of a single KPI-Attribute relationship (one to one). A KPI attribute may appear in many KPIs (zero, one or more).
- KPI-to-SLA. SLA may reference any number of KPIs or none, if it consists of custom SQL (zero, one or more). A KPI may appear in many SLAs (zero, one or more).
- SLA-Group-to-SLA. SLA group may contain any number of SLAs (zero, one or more). An SLA must be categorized under one SLA Group (one to one).
