Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
CAN/CSA-ISO/IEC 10164-11-97: Technical Analysis of the OSI Workload Management Function
Scope, Requirements, and Compliance for Systems Management using ITU-T X.739
1. Scope and Fundamental Context of CAN/CSA-ISO/IEC 10164-11-97
This article provides a detailed technical analysis of CAN/CSA-ISO/IEC 10164-11-97, the Canadian adoption of the international standard Information technology — Open Systems Interconnection — Systems management: Workload management function. This standard is technically identical to ITU-T Recommendation X.739. Ratified by the Standards Council of Canada in 1997, it defines a critical component of the OSI Systems Management framework (ISO/IEC 7498-4).
The primary scope of this standard is to specify the semantics and behavior of a Systems Management Function (SMF) used for monitoring and controlling the workload of managed objects within an Open Systems Interconnection (OSI) environment. It addresses the need for standardized mechanisms to evaluate system performance, detect anomalies, and plan for capacity across heterogeneous network elements. The standard defines an abstract model of workload management that is independent of any specific implementation or protocol stack, although it is specifically designed to operate over the Common Management Information Protocol (CMIP).
Standard Alignment: CAN/CSA-ISO/IEC 10164-11-97 is fully aligned with the global ITU-T X.739 specification. Organizations looking for a multi-vendor, standardized method for workload monitoring can use this as a foundation for OSI-based management domains.
The standard defines the workload of a managed object as the measure of its current usage of resources relative to its capacity. This function provides a set of services to establish workload thresholds, receive threshold crossing alerts (i.e., workload reports), and query the current workload value of a specific managed object.
2. Core Technical Requirements and Managed Object Model
The technical backbone of CAN/CSA-ISO/IEC 10164-11-97 lies in its well-defined Managed Object (MO) classes and their associated packages, attributes, and notifications. The standard uses the Guidelines for the Definition of Managed Objects (GDMO) to specify these components.
Key Managed Object Classes
The standard introduces several specific classes to implement the workload management function. The central entity is the workloadMonitor managed object.
Table 1: Primary Managed Object Classes
| Managed Object Class | Description | Key Attributes |
|---|---|---|
workloadMonitor | Responsible for monitoring the workload of a specific managed object. It holds the current workload value and the associated threshold definitions. | workloadThresholdHigh, workloadThresholdLow, workloadHysteresis, currentWorkloadValue, observingManagedObject |
metricObject | Represents the underlying resource or counter being measured (e.g., CPU utilization, queue length, process count). | metricId, metricValue, metricModTime |
thresholdData | Defines the specific thresholds (high/low) and the parameters for the type of workload measurement. | granPeriod, thresholdInfoId, measurementDuration |
workloadReporting | A conditional package that allows the system to emit an M-EVENT-REPORT when a threshold is crossed or cleared. | workloadReportId, notificationDiscriminator |
Services and Operations
The standard mandates the use of specific CMIS primitives for interaction. An administrator can use M-GET to retrieve the current workload value from a workloadMonitor object. M-SET is used to modify threshold values dynamically. M-EVENT-REPORT is used by the agent to inform the manager of workload anomaly events (Threshold Crossing Alerts – TCAs).
Implementation Criticality: The
workloadHysteresis attribute is crucial for preventing “event storms”. The standard allows hysteresis to be defined as either an offset value (the workload must deviate by a specific amount from the threshold before a new event is generated) or a time-based value (the workload must remain outside the threshold for a specified duration). Implementers must correctly handle this dual nature to ensure stable behavior. 3. Implementation Highlights and Operational Mechanisms
Deploying a system that conforms to CAN/CSA-ISO/IEC 10164-11-97 requires a robust OSI management stack. The workload management function is typically implemented in an agent system that monitors a network element or a software process.
Workload Monitoring Lifecycle
- Instantiation: A
workloadMonitorobject is created and associated with a target managed object via theobservingManagedObjectattribute. - Configuration: Thresholds (
workloadThresholdHigh,workloadThresholdLow) and the hysteresis value are set by the managing system. - Operational Monitoring: The agent continuously assesses the workload of the target object. If the value exceeds the high threshold or falls below the low threshold, the system checks the hysteresis before acting.
- Reporting: If the threshold condition remains stable for the duration defined by the hysteresis, an
M-EVENT-REPORTis sent encapsulating the relevant workload data.
Table 2: Workload Monitoring Parameters (as defined in X.739/10164-11)
| Parameter | Type | Definition |
|---|---|---|
| Workload Value | Dynamic | The current metric representing the system load (e.g., percentage of CPU in use, queue depth). |
| Granularity Period | Duration | The interval over which the workload value is calculated or sampled (e.g., 5 seconds, 1 minute). |
| Threshold High | Integer | The upper boundary value. Crossing this from below triggers a highWorkload report. |
| Threshold Low | Integer | The lower boundary value. Crossing this from above triggers a lowWorkload report. |
| Hysteresis | Integer or Duration | A dead-band value. It defines either the offset by which the workload must exceed the threshold before the condition is true, or the time it must remain outside the threshold. |
Best Practice for Managers: When polling workload data using
M-GET, request the currentWorkloadValue and the workloadThresholdHigh/Low attributes together. This provides instant context on whether the current load is in a normal, warning, or critical state without requiring separate requests, reducing management traffic. 4. Compliance, Audit, and Conformance Notes
Conformance to CAN/CSA-ISO/IEC 10164-11-97 is vital for interoperability in strict OSI/CMIP-based management environments. The standard provides a PICS (Protocol Implementation Conformance Statement) proforma (usually found in an informative annex) that defines the capabilities and options an implementation must declare.
Conformance Requirements
- Mandatory Support: An implementation must support the
workloadMonitorclass, including theworkloadMonitorPackageand themetricObjectPackage. - Conditional Support: The
workloadReportingpackage is conditional. If the agent supports reporting, it must fully implement theworkloadReportnotification and the associated discriminator constructs. - GDMO Compliance: The ASN.1 definitions for the attributes and notifications must be compiled correctly. The naming hierarchy of the
workloadMonitorobject must be consistent with the name binding specified in the standard.
Common Non-Compliance Issue: Failing to handle the
granPeriod and measurementDuration attributes correctly is a frequent source of certification failure. The managing system expects the values reported to represent the average load over the defined period, not an instantaneous reading. Misreporting this leads to false positives in threshold crossing alerts and invalid capacity planning data. The standard also links strongly to other OSI management functions. It must be used in conjunction with ISO/IEC 10164-1 (Object Management Function) for creating and deleting the workloadMonitor instances, and ISO/IEC 10164-5 (Event Report Management Function) for configuring the forwarding of workload reports to different manager destinations.
Strategic Value: For organizations maintaining legacy OSI systems (e.g., Telecommunications Management Network or specific defense networks), rigorous adherence to CAN/CSA-ISO/IEC 10164-11-97 ensures that workload management is handled in a predictable, standard-compliant manner, facilitating seamless integration between managing and managed systems from different vendors.
Frequently Asked Questions (FAQs)
Q: What is the main difference between CAN/CSA-ISO/IEC 10164-11-97 and the original ISO/IEC 10164-11:1995?
A: The difference is purely administrative in the context of adoption. CAN/CSA-ISO/IEC 10164-11-97 is the identical adoption of the international standard by the Standards Council of Canada. The technical content, managed object definitions, and services are exactly the same as the ISO and ITU-T versions. The 1997 date denotes the year of Canadian ratification.
A: The difference is purely administrative in the context of adoption. CAN/CSA-ISO/IEC 10164-11-97 is the identical adoption of the international standard by the Standards Council of Canada. The technical content, managed object definitions, and services are exactly the same as the ISO and ITU-T versions. The 1997 date denotes the year of Canadian ratification.
Q: Is CMIP strictly required for implementing this Workload Management function?
A: Yes, this standard was designed within the OSI Systems Management framework, which relies on CMIS/CMIP services for interaction. While the abstract model and GDMO definitions could conceptually be mapped to other protocols like SNMP (via a proxy agent), a native, conformant implementation must support the CMIS primitives (
A: Yes, this standard was designed within the OSI Systems Management framework, which relies on CMIS/CMIP services for interaction. While the abstract model and GDMO definitions could conceptually be mapped to other protocols like SNMP (via a proxy agent), a native, conformant implementation must support the CMIS primitives (
M-GET, M-SET, M-EVENT-REPORT). Q: How does the standard handle historical workload data?
A: The primary focus is on real-time monitoring and threshold crossing alerts. However, the standard defines mechanisms to capture workload values over a granularity period (
A: The primary focus is on real-time monitoring and threshold crossing alerts. However, the standard defines mechanisms to capture workload values over a granularity period (
granPeriod). The management system (manager) is responsible for logging the received currentWorkloadValue or workloadReport events into its own database for historical analysis and capacity planning. The agent itself does not inherently provide a long-term data store as part of the workload function. Q: Where can I find the GDMO definitions for the
A: The complete GDMO definitions, along with the ASN.1 module for the workload management function, are provided in the normative annexes of CAN/CSA-ISO/IEC 10164-11-97.
workloadMonitor class? A: The complete GDMO definitions, along with the ASN.1 module for the workload management function, are provided in the normative annexes of CAN/CSA-ISO/IEC 10164-11-97.