DCIM: what are the complementarities with ITSM practice? (Part 5)
By Laurent Duenas, December, 2nd 2013
Part 4 presented the intersections between DCIM and ITSM about industrialization, operations, and change management. This chapter continues this comparison over other perspectives such as availability, capacity, asset management.
What are the links between DCIM and ITSM? (suite)
Improve the availability of the infrastructure
Availability management is an essential process that aims to achieve the availability objectives & targets for It Services as expected by IT department’s customers. These objectives are mainly guided by the cost of unavailability for the company. According to Gartner Group studies, the Data Center unavailability cost is evaluated between 100 K$ by hour for the smallest ones and up to 6000 K$ by hour for the biggest.
Availability Management activities can be summarized as follow:
► Monitoring the IT components in order to anticipate any risk of unavailability,
► Analyzing the cause of unavailability and/or identifying the single point of failure,
► Defining Norms & Standards that improve availability*
► Verifying that Norms & Standards applied improve availability (as expected).
* 2 kinds of norms exist: the ones which increase the component and IT services resilience - they are called “design for availability -; the ones that facilitate the recovery of an IT Service following an incident -they are called “design for recovery”).
Here again, DCIM solutions provide essential contributions to Availability Management activities. Monitoring functions and simulation-for-analysis functions bring all required means to investigate the cause of unavailability and achieve improvement objectives.
Examples: the analysis of failure rate of UPS or transmission equipments helps to improve the resilience of power supply circuits. Analysis of virtual machine flows from ones blade racks to others helps to anticipate heating concentration and identify weaknesses in infrastructure architecture. DCIM simulation tools contribute to prepare moving or reorganization of the Data-Center and find the best distribution of machines to avoid mix of hot and cold-air flows resulting to cooling energy waste.
Taken from a global perspective, resilience principles and design for recovery, which are normally defined at IT Service level (in the ITSM sphere of management) should be declined into all technological domains. Within a same component chain – delivering an IT Service – the global resilience obtained is aligned onto the availability of the weakest component. It is essential that DCIM tools provide the information of physical infrastructure to support availability policies. DCIM tools’ mission should be to keep Infrastructure availability objective on-track as defined in the ITSM sphere.
Contribute to Asset Management
Monitoring of machine aging, measuring failure rate; producing warning or utilization rate to engage replacement campaigns or preventive maintenance; enabling accurate localization of physical components in order to secure maintenance operations; are all functions provided by DICM tools which are assimilated to Asset Management.
Example : DCIM tools trigger proactive maintenance visits on cooling or power supply units after an alert is being issued or after a level of utilization within a period of time.
It is also correct to consider that Asset Management activities are tightly linked to the component nature. Therefore, rules which apply to Data Center equipments do not obviously work for other IT components. A sort of independence with Asset Management relating to other IT domains can be concluded. However, objectives and policies of Asset Management find their origins into the ITSM* processes, and they do work for Data Center too :
► Service Asset and Configuration Management (SACM) for inventory and audit policy,
► Change Management, Release Management for planning and performing IMAC (Install, Move, Add, Change),
► Availability Management for replacement campaign for IT configurations aiming at deploying a more resilient architecture,
► Financial Management for structuring and consolidating financial information, providing relevancy for budget decisions,
* The notion of Asset Management as it is employed in the organizations is not identified as a unique process as it is for the Service Asset and Configuration Management process, at the opposite of what it should be thought by a majority of people. The management of the repository (the asset data base) is part of the Service Asset and Configuration Management process. The update of the data base as inventory and audit tasks are ones of its activities. The management operations executed on components (and assets) take place in several ITSM management processes, according to their objectives, such as changes (Change Management), deployments (Release and Deployment Management), retrieval (Release and Deployment Management), failure rates (Incident Management and Problem Management), etc.
For same reasons previously discussed, the necessity of aligning Data Center to ITSM for Asset Management objectives is a reality. Due to the difference of nature between Data Center equipments and other IT components, ITAM* alignment mostly applies at a rule and policy level rather than at an operation level.
* ITAM for IT Asset Management
To anticipate capacity needs
Modelize IT resources consumptions through utilization profiles, as ITSM could imagine it when talking about Service Capacity Management, has a tight link with DCIM capabilities as well. Even if today, concepts are not all applied and shared, such as “User profiles” and “Pattern of Business Activity”, DCIM tools provide consolidation features for physical resources capacity and measurement through consumption units.
For example, it is possible to gather capacity information of a set of physical equipments contributing to a specific energy flow. By consolidating this information, it can simulate global workload capacity, and anticipate redundancy and security cutout for a specific area (as for a range of bay, for one aisle, for several aisles, for a full room, etc). It can also mix different energy capacities, such as power supply, air cooling, and protection, for a specific location/place in the Data Center, or from a more abstractive perspective, such as a class of service. This aims at anticipating hosting capacities for more end-to-end infrastructure services at a Gold, Silver, or Bronze level.
According to the Capacity Management activity classification used in the ITIL V3 framework, DCIM tools are positionned at the “Resource Capacity Management”, because they measure, simulate and plan consumptions at physical resource level. But, they move closer to Service Capacity Management level when they aggregate consumptions at a superior unit level. This overpasses the technical angle of the unitary resource. It gives an indication closer to business usage and increases predictability of all components of the chain.
Of course, all DCIM tools do not give a full vision of the ITSM scope (end-to-end vision of the IT Service), but they do it in their proper way by consolidating at Data Center infrastructure level. Hence, all functions of end-user profiling or consumption calculation based on IT infrastructure Service consumption unit contributes to have Service Capacity Management level, providing a closer vision to reality improving risks and requirements anticipation.
The facts are, that despite the possibilities provided by these tools, Capacity Management process still be fragmented by technological domains. You would hardly find transverse processes federating common or even consistent policies, for all technology domains, based on an IT Service oriented vision. Each actors playing in the Capacity Management field keep specialized into their own competence domains. This brings methodologies, objectives, and even more due to different specialized tools, to silos. This is true for pure IT components, such as Storage cabins only managed by storage system managers, logical Databases managed by DBA, Network bandwidth managed by Network engineers, etc. Service Capacity Management working in 2013 transversally on all components contributing to the delivery of a same IT, is far from being a reality.
Bridges between IT Service consumption models (as seen in ITSM practice) and those employed for Data Center infrastructure (managed by DCIM tools) must be tightly built in order to take advantage of the valuable information on Service capacities.
Copyright © 2013 - PRACT Publishing
► No link
► No link
► No product