Methodology for Selecting Data Center Design Class Utilizing Performance Criteria

Posted on March 29, 2012 by

BICSI Supplemental Information 001 is a document freely available that addresses a revision to the ANSI/BICSI 002-2011 Data Center standard.

I submitted a revision to the published methodology in Annex B “Reliability and Availability” to incorporate the corrected three step analysis for determining the appropriate level of redundancy required for a data center, which includes;
 
1.) Operational Requirements (available planned downtime)
2.) Availability Requirements (unplanned downtime objectives)
3.) Impact of Downtime

 

The data center standards committee reviewed the submitted revision and approved the “Methodology for Selecting Data Center Design Class Utilizing Performance Criteria” supplemental information on February 12, 2012. The data center redundancy classification and methodology to identify the appropriate level of redundancy is freely available to download and distribute at Isaak Technologies website.

Also, there was a revision to the “Impact of Downtime” terms used within the standard. These revised terms were contributed by participants in one of the DC125 data center design courses that I instructed.

Posted in ANSI/BICSI 002-2011, Data Center Reliability, Isaak Technologies | Leave a comment

Data Center Design: End-to-End Reliability

Posted on November 26, 2011 by

I recently finished writing a white paper on the Data Center Design End-to-End Reliability, covering the content that I presented at Interop New York.  The content was very well received and generated discussions that extended into the corridors after the presentation was completed.  The 45 page white paper is available on Isaak Technologies website.

The core objective of the end-to-end reliability alignment is to ensure that all facility and IT systems are designed and configured in a methodical approach using a common framework to define the reliability design criteria.

Posted in Data Center Conference, Data Center Design, Data Center Reliability, Isaak Technologies | Leave a comment

Data Center Design – How to Align Services Reliability Discussed @ Interop New York

Posted on September 15, 2011 by

Isaak Technologies, a data center design firm, will be presenting at Interop New York 2011, happening October 3-7, 2011 at the Javits Convention Center.

I will discuss methodologies that will help guide managers and executives in their decisions regarding design and operational requirements for data center services; whether in-house data centers, outsourced colo or managed data center services, or cloud based services. The recently released ANSI/BICSI 002-2011 Data Center Design and Implementation Best Practices standard is one tool that can be used to help determine specific requirements.  The “performance-based” approach of this ANSI standard will be reviewed,  as well as how it can help guide decisions regarding operational requirements or design reliability and redundancy requirements for the data center facility and the IT enterprise architecture.

The discussion will focus on:

  • How to identify appropriate reliability criteria, based on the user’s specific business objectives, for the data center services; facility and IT enterprise architecture.
  • How to apply consistent methodology in reliability analysis for the data center services across the enterprise
  • How to strategically implement applications within a multi data center enterprise with a combination of in-house data centers, outsourced colo or managed data center services or cloud based services.

About Interop ®
Interop® drives the adoption of technology, providing knowledge and insight to help IT and corporate decision-makers achieve business success. Part of UBM TechWeb’s family of global brands, Interop is the leading business technology event series. Through in-depth educational programs, workshops, real-world demonstrations and live technology implementations in its unique InteropNet program, Interop provides the forum for the most powerful innovations and solutions the industry has to offer. For more information about these events visit, http://www.interop.com/

Posted in ANSI/BICSI 002-2011, Data Center Conference, Data Center Design, Data Center Reliability, Isaak Technologies | Leave a comment

Latin America Data Center Summit – San Jose, Costa Rica

Posted on July 26, 2011 by

Since I have been asked to speak at the Latin America Data Center Summit at the end of August in San Jose Costa Rica on the new ANSI/BICSI 002-2011 Data Center Design and Implementation Best Practices standard, one of my main questions is who will end up being the predominant audience of this standard? The data center standard was written for the following audiences:

  • Information Technology
    • Designers
    • Telecommunications/Network Management
    • Operations Management
    • Security Management
  • Facilities
    • Architectural/Engineering Designers
    • Facilities Management
    • Physical Security Management

If attendance at recent BICSI DC125 Data Center Design courses is any indication it may roughly follow 44% end users/data center managers, 22% consultants, 22% contractors, and 12% venders/manufacturers.  Since past course attendees have all felt they recieved value from the content it would appear that the standard may appeal to all stakeholders within data center design, management and operations.

Posted in ANSI/BICSI 002-2011, Data Center Conference, Data Center Design, Isaak Technologies | Leave a comment

Power Outlet Units (POU) – Managing Overload Conditions

Posted on May 19, 2011 by

The other day I had a client ask me what they should be paying attention to when monitoring the load on the 208V/3Φ 30A POUs within their server cabinets.  The client had been instructed by the POU vendor that as long as the local meter amperage reading on the POU did not exceed 80% of the supply circuit breaker rating feeding the POU he would be fine.

If all the IT loads are equally balanced across all three phases then this would be true.  However, loads that are un-balanced across the phases may overload one of the receptacle bank’s protection integrated into the POU, while the overall POU load is still below 80% of the circuit breaker rating feeding the POU.

To demonstrate this, let’s use an example of 9 servers connected to a POU.  These loads would typically have dual power supplies and be connected to an “A” and a “B” POU, but for the purpose of this example we will assume all the IT equipment is powered from one POU source (which is a fail mode that should be planned for anyways).  The POU and 9 servers consist of:

  • (3) servers with an actual load of 1000VA & C19 power connections
  • (6) servers with an actual load of 530VA & C13 power connections
  • The POU is feed from a 208V/3Φ 30A circuit (80% rated)
  • The POU has integral 20A breakers or fuses, 100% rated, for all 208V/1Φ line-to-line circuits
  • Total POU load is 6180VA, well below 80% of the rating of the circuit breaker rating feeding the POU (208V x 30A x √3 x 0.8 = 8646VA)

 

All of the IT loads are evenly distributed across all the source phases in the configuration shown above.  No problems with the integral 20A breakers/fuses or the 30A supply circuit breaker.

The IT loads are distributed across the top two receptacle banks in the configuration shown above.  No problems with the integral 20A breakers/fuses, but one phase of the 30A supply circuit breaker is in an overload condition (greater than the 80% rating) and is susceptible to tripping the supply circuit.

The 1000VA IT loads are distributed across all the source phases, but the 530VA IT loads are connected from the top down all within the top receptacle bank in the configuration shown above.  No problems with the 30A supply circuit breaker, but the integral 20A breaker/fuse for the top receptacle bank  is in an overload condition and is susceptible to tripping the top receptacle bank integral breaker/fuse.

The local meter amperage reading for the POU this particular client was using was measuring the Phase A, Phase B and Phase C supply current.  As you can see it is also important to pay attention to the amperage of the combined IT loads connected to each of the three receptacle banks on each POU.   The client was able to easily manage these loads by monitoring the receptacle bank amperages using the power management tool they integrated with their POUs.

Posted in Data Center Design, Data Center Power, Isaak Technologies | Leave a comment