Why fine tuning a network can be bad? ( post #1)

Hello Everyone,

Recently I posed a document to my boss suggesting the outline for creating a Best Practices guide for our network, and two issues came up.

Here is the Outline I created:
IT Department Best Practices Outline v0.2

Proposed core topics which *could* be contained in a Best Practices volume:

• The organizational purpose of network.
• Network Architecture design and implementation procedures.
o Expansion needs both current and in the future.
o Network quantification and measurement
o Network Topology assessment and implementation.
• Governance for acceptable use.
o Policies and procedures for network users
o Quantity of both users and computers on the network.
• Security
• Compatibility issues
• Hardware/Software Procurement
o Asset management and tracking
• Disaster recovery and management (see DRP)
• Fine tuning NB:


The first issue he (My Boss) had was this:


Boss: I would like you to start addressing the "Governance for acceptable use" section.

One issue that we have experienced is developers sharing passwords, who knows the root password to which server, etc. That needs to be addressed.

(This was addressed already in this forum.
One of the solutions is to use "sudo" to give the users root but also log them.

On the issue of "Fine Tuning" What has worked for you in this arena?


Boss: We do not do fine tuning! write me a one paragraph essay on why we do not do any fine tuning.

To this I have no clue (well, SORT of a clue. -see next post) But I am not 100% certain if I am on the same page as my boss.
Doing a websearch reveals nothing on the topic of why this would be a bad thing. (Fine tuning, that is)

I thank you in advance for your wisdom on this matter.
(Just a point in the right direction would be helpful)

Ok, here are my thoughts on the subject:

Once the network is running, and stable, one can fine tune speed (Unless the replacement of upgraded harware imposes too much of a risk of downtime, or service interuptions.) In which case, defining a policy on upgrade procedure would also be a good idea.
But moreover, "Fine-tuning" my be defined at least in my eyes as tweaking the system to provide better performance from a software level.
This *could* cause extra issues to pop up I believe.

Is the axiom "If it aint broke, don't fix it" applicable?
What do you think?
Perhaps my boss is trying to tell me this?
What could I be missing here?

After some further examination of the subject, I have so far this platform as to why my Boss took me to task on the issue of "Fine-Tuning" the network:

The necessity of fine-tuning may lead to various problems that do not show that our theories are incorrect, in the sense of disagreeing with our observations, but nevertheless indicate that we have overlooked something. In setting up a network correctly from the onset, and carefully planning and addressing all seen issues and allowing for unforeseen problems during the process would negate the necessity for “fine-tuning” at all.

In other words, Do it once, do it right.


What do you think?

Here is my next to final draft:

Why “Fine-Tuning” in a Network Environment does not exist in our verbiage.

The necessity of fine-tuning may lead to various problems that do not show that our planning and facilitation of the designed network is incorrect, in the sense of disagreeing with our schema, but nevertheless indicate that we would have overlooked something, which clearly exposes unnecessary remissness.
In pursuing development of a network correctly from its onset, and by carefully planning and addressing all known issues, and creating controlled allowance parameters for unforeseen problems as part of the provisioning process would negate the necessity for a “fine-tuning” the resulting product at all.

In other words, proper planning phase is core and supersedes execution phases in proactive issue elimination
I.E., “Do it once, do it right.”

In conclusion: We do not “fine-tune” per say, rather we can modify within pre-outlined events as ongoing situations (which are also anticipated) dictate.

~Hopefully my Boss was looking for this line of thought and not: "42".

Well,
The results are in, and sad to say I missed the mark entirely.
Which is not to say that my previous essay was not valid, however, my Boss has informed me that the reason why we do not "fine tune" a network is because it is the same as doing nothing at all, and of course "doing nothing" is not what we are paid for.

"Fine Tuning" is not measurable to the customer, it produces no results that can be immediately seen, and in some cases, too much fine tuning can break things over time. We *always* address our efforts on issues which have high impact, are of importance, and which have results that are visiable to the customer and non-technical management.

So, the ending result of my journey for today was:
Fine tuning does not add value to the business.

Fine Tuning = not addressing important issues = Doing Nothing = unemployment line.

Have an excellent weekend!
Be well!