Operations
I have two sessions for VMworld 2009: one on Capacity Management, and one on High Performance Operations – both are based upon, of course, vSphere..
This page is the home for my collections around High Performance Operations and will be the source for both the presentation and the companion material. My intention is to share as much material as possible prior to VMworld 2009 to find an audience, and then keep the discussion going after VMworld 2009 via this site (and others). The presentation at VMworld is only 45 minutes, and I need to get the following three points across:
- think about operations differently – the difference between ok-ops and hi-ops is in here (points to head). Remember that the GM and Toyota have the same machines, but one is inefficient and ineffective and the other is Toyota – the difference is in the mindset, culture and approach, and IT is no different.
- list actions that differentiate ok from high – efficient means codifying successful ways, and constantly evolving the process – not sticking to static stuffy steps.
- see what is possible – thinking and acting in a new way – here’s what’s possible with a case study
My naked goal is to elevate Operations out of the mundane and neglected corner of enterprise IT and into the spotlight, front and center. The following pages are part of this goal:
Why isn’t Operations driving design decisions? Instead, it receives a badly made ball over-the-wall from the engineering team.
Why reduce the amount of investment in Operations, when the opposite will actually increase efficiency and effectiveness if done right?
I spent five years at VMware working on Operational Readiness and seeing with my own eyes the difference that great operations makes to virtualization: it can mean a difference in guest:host ratios of 3:1 in an ok-ops environment, to 20:1 in a great ops environment. Other studies have shown that high performance operations increases throughput (new projects) by eight times. Imagine that impact on the business if you are a bank: eight times faster than your competitors; this is the difference great operations can make. Not to mention customer satisfaction, availability, disaster recovery and the fact that when Operations is done great, everyone loves their job
Hi I think your point is very valid and as a Vmware consultant I see this problem every day: Purchase and design decisions are made based on assumptions, managers bias for certain vendors or just emotional driven considerations. There is no better way to take good decisions that based them on “numbers”, historic data has a tremendous value, not only for trending purposes also to understand where are the gaps to match the magic 20:1 ratio (or whatever number) was establishes during the virtualization assessment phase.
It seems that the famous technology-process-people concept when companies virtualize environments is in terms of time and effor: 90% technology, 5% process, 5% people.
Andres
twitter: @virtualizemos
I think all architects should be required to come up through operations and then engineering. A lesson i learned during work experience in my Uni years. We student Aero engineers had to go work with aircraft mechanics/builders/repairers. We were told, “your designs are worthless if these guys cant work with them”. I was shown an example, “see this rivet up inside this cavity here, how the hell am i meant to get that out? i haven’t got 6 foot long arms!”
Put your designs in front of Operations people, they’ll find the weaknesses. Think of it as fitness training for your design skills.
Another good saying, “an ounce of prevention is worth a pound of cure”