There is an updated version of this document available on vPivot. Please refer to that document for the newest version of the ladder.
My new responsibilities in the Asia Pacific and Japan (APJ) region have put me in front of a great deal of customers recently. With the exception of Australia and New Zealand (ANZ), APJ is newer to virtualization yet just as eager to virtualize fast. Customers, partners, and colleagues have asked me how they can measure their team members’ VMware proficiency and what can be done to increase skills. Today I am going to present my first pass at the measurement part of this problem and I ask for your comments on my system.
My goal is to provide an objective system by which everyone in the virtualization community can identify their relative strength among their peers. The names and descriptions are my own. The indicators can be used as a hint to an evaluator that a person has reached a certain level of the ladder. The measurements are a rough cut at tests that could be applied to anyone that wishes to demonstrate they have reached a certain level of proficiency. I hope to build up these tests’ complexity over the coming weeks.
Without further ado, here is my first effort at the SBD VMware proficiency ladder. I request your comments below.
| Name | Description | Indicators | Measurements |
|---|---|---|---|
| Uninformed | Has not used VMware products. | Unfamiliar with terminology, no experience with virtual machines. | Not applicable. You do not need to pass a test to be labeled ‘uninformed’. :) |
| Novice | Rudimentary understanding of VMware’s work. Not ready for VMware administration responsibilities. | Has played with virtualization products, usually as a hobbyist or in support of other professional activities. Knows how to build a virtual machine. Knows how to install products. May not know major components of VMware Tools and can likely not name management products. | Given ISO and a hosted product, can get two virtual machines running and on the network. |
| Proficient | Administrator-level familiarity with vSphere or VI3. | Has used VMware’s enterprise products and knows the purpose of most of the management tools and has used at least a few of them. May not have designed a large infrastructure and missing experience with some management tools. | Has VCP. Can configure a vCenter-managed cluster of ESX hosts and enable HA, DRS, and FT. Can describe the purpose of the components in VMware Tools. Can configure the vMA and show important counters in resxtop. |
| Advanced | Experienced VMware administrator. Could support a large deployment. | Knows all of VMware’s products, has played with all of them, and may have implemented many in a production environment. A fair understanding of operating systems, hardware infrastructure, and security. Can recognize problems with operational readiness and suggest corrections. | May have VCAP*. Can demonstrate installation and configuration of all management tools. Can configure the Nexus 1000V. Can build a View environment. Can correct performance problems using multiple tools. |
| Expert | Capable of succeeding as a virtualization datacenter architect for a moderate to very large datacenter. | Has understanding of product internals, management tool weakness, and troubleshooting techniques for configuration or performance problems. For any environment, familiar with hardware (server configurations, network design, storage design), software (operating systems and applications), and process (disaster recovery, backup strategy, security, change management). Is following the virtualization community and may be participating regularly. | May have VCDX. Can give a five minute whiteboard session on the value, strengths, and weaknesses of any product VMware sells. Can lead a discussion on operational readiness and measure a company against key criteria. Can name several virtualization blogs and the last few important entries on each. |
(*) Unlike the VCP and VCDX, I do not have much familiarity with the rigor of the VCAP program. I offer this benchmark as a possibility and encourage your comments.
I am plan on toying with this model for several weeks. I would like your help with this. If you do not already have a long list of suggestions, consider the following:
- Would it be considered an asset or a sign of experience that someone has engaged VMware support?
- Should a knowledge of VMware licensing fall into this ladder or is it merely optional for the job of virtual infrastructure management?
- Could scripting with the Power CLI or SDK be used as a measurement against higher levels or experience with scripting reasonably be considered an indicator for higher levels?
- Is four levels (excluding the “uninformed” state) the right number? Three is the minimum possible, I think. I fear that too many rungs make it difficult to unambiguously assign someone to a level.
I look forward to your comments.
I wouldn’t put contacting support as a negative or positive. Personally I never have because I like figuring things out on my own but I can see certain situations where you might especially if it was related to a bug. Now if someone was contacting support for every little thing I would put that as a negative because they are taking the lazy way out.
I think every VMware admin should be familiar with all the different licensing options.
Some people are good with scripting and some are not, it requires a different mind set and skill set than administration. I guess I could see it as a plus, it’s helpful in some environments and not so much in others. But I don’t think I would hold it against someone if they couldn’t script.
There are experts and then there are experts, I can definitely see a few being above the regular expert level. Maybe add a God level.
As per your ‘ladder’ description, an alternative to shoe-boxing individuals would be to come up with a matrix composed of points.
This will allow you to not only make the model a dynamic one, as you could extend it by simply adding a question and an associated point-score; but also more indicative of the value the person brings to VMware-knowledge-table.
You could even apply a weight system where certifications give you 10-15 points, yet individual bash/python scripting skills count for 1-2 points.
One important consideration to make is that not everyone that is very proficient in datacenter design+admin has touched VMware View. Every deployment doesn’t always have a VDI need; tying and marking merely by products with disparate markets segments will dilute your ladder.
In theory VMware certifications should be a measure of proficiency at three different levels of proficient, advanced, expert: VCP, VCAP, VCDX. The blueprints for those certs would map to what those levels should know.
Not sure if would include Nexus 1000V. I would regard myself as at least Adv, if not Expert – but I’ve not touch Nexus 1000V. It’s not through lack of interest, but lack access to the technology, and a lack of demand for it from my mainly HP customers. As for DCAP – going forward fine, but right now these exams aren’t even available just yet… I would like to see a recognition of a history – I’m VCP2/3/4 which demonstrates maturity with the product beyond the vSphere4 release…