Labeling CI

[mlopezluque]

Hello, we are looking for a ITIL compliant system to label our inventory. Could someone give us some example.

Thanks in advance.

[BorisBear]

Yes, I'm sure someone could.
_________________
ITIL Lilac Badge

[UrgentJensen]

Hi mlopezluque,

You can't be 'compliant' to ITIL so if you ask for examples, don't expect an industry standard.

UJ
_________________
Did I just say that out loud?

(Beige badge)

[UKVIKING]

mlopezluque

There is no such thing

ITIL is a best practice for people, processes etc

ITIL can not be compliant, approved, certified (except for individuals).

Tools etc can be process focused - incident, problem, etc

but they can not be compliant
_________________
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter

[UrgentJensen]

Best thing to do is make it up, anything, then when someone asks the same question on here in a couple of weeks you can give your suggestion and then we'll have a open framework.

Sorted.

UJ
_________________
Did I just say that out loud?

(Beige badge)

[mlopezluque]

Hello, I only was looking for how people is labeling inventory out there. Now we are labeling PCs whit the username and when we change a user PC we change the PC name to use the same username. With autodiscovery tools and using this method we lost the history for the PC. We are thinking in labeling the PCs using PC000001, PC000002, ... and we wanted to know is this could a good method.

Thanks in advance.

[UrgentJensen]

Oh well that's a simpler question. Again there is no right or wrong as it depends on what suits you, but you could go with something like this:

LDNSLSXP1234

Where LDN is a three letter abbreviation of the pc location e.g. London.

Where SLS is a three letter abbreviation of the deparment.

Where XP is the OS.

Where 1234 is a unique number for the pc in that department.

This makes it very easy to work out what machine or other device you're looking at.

Hope that helps.

UJ
_________________
Did I just say that out loud?

(Beige badge)

[UKVIKING]

Or you can use the following

srvr for server
wks for workstation (PCs)
lap for well laptops
pda for well pda
web for web servers
smtp for mail server

etc etc

rtr 72 uk 01

rtr router
72 cisco 7200 series router
in the UK

01 first router

or you can do things like this

routers named after tool companys - makita, bosch
distributed directors after movie directors
content distributors after detectives
etc

this of course is ok for a small company but does not scale well


Your company needs to define a naming convention that is usefule to you
_________________
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter

[milligna]

This is a tough concept. In one place I worked at we used to label PC CIs with the hostname of the PC as defined by the desktop team. Of course as soon as the PC was reimaged with a different hostname, a new CI was created (we had discovery tools which were auto-creating CIs). We also encountered issues when PCs where created with the same hostname (after the previous PC had been decomissioned). My suggestion would be to have a unique name (maybe hostname) along with a physical unique identifier (serial number). I the example I gave this would have taken care of MOST of the problems we encountered (there were still times where a PC motherboard was swapped out and the autodiscovered Serial number gets changed - hence changing the discovered Serial number).

[pel]

We use an extremely simple approach.
All servers are labeled
M<uid> - <uid> is a unique number (no letters)

The idea is that it is not much use to label with anything very informational.
E.g. both the name and DNS-name of the server might change many times during it's life time. That's true for operating systems to.

We choose this simple approach since we need to keep track of the hardware and we wanted it to be simple to remember the name. All the meta information is stored in the CMDB anyway - so that is useless (and subject to change).

[UrgentJensen]

Pel, milligna,

Environments with large numbers of desktops and laptops, often distributed over various geographic sites benefit from a clear naming convention.

Configuration Management is still an immature area, but there is clear value-add from having desktop support teams who can easily identify machines and resolve incidents quickly.

I think we can all agree though that it's down to what's right for the organisation you work in and the processes you need to have in place.

UJ
_________________
Did I just say that out loud?

(Beige badge)

[UKVIKING]

Millingna

The true answer is 'it depends'

Things to consider

Any device on the Intra/Internet has the following

Hostname
Fully Qualilfied Domain name
IP address / Subnet / Gateway IP

these can change

certain pieces of information are absolute

MAC addresses on Network cards dont change. From what I recall, each vendor gets a range. The odds of get 2 identical MAC address is ////\\\

Device serial number can change when the chassis is replaced.

Most company have Unique Company Asset Tag IDs.

The combination of the

Asset Tag ID
IP data set
hostname

does create a unique ID set for a device.

For example: if the PCs are being tracked / managed, there should be the following details - ACME01000101. 10.1.1.2, mypc

Regardless of what internal parts are being switched out. That PC will always be that PC - UNTIL IT IS REPLACED as a whole unit.

You should only go to the smallest complete package that you can reasonably track. There is no sense in spending pounds, euros or dollars in time and money to track pence, cents.

The lowest CI for front end customer / user devices is the PC - Laptop, desktop, PDA, Blackberry.

If the devices are controlled via the corporate asset tag, then there is the primary ID, the secondary ID is the hostname & IP address (in DHCP world, the IP changes over time but may be in a specific range)
_________________
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter

[milligna]

Hi John,

Traditionally for desktop PCs in most cases the Operating System's configuration is going to be intrinsically linked to the physical device, and these are not going to change very often, however times are changing and I see this as not being true for much longer.

We're already seeing this with high availability virtualisation in the server space - My preference for Virtual servers is to track each OS instance separately from the hardware... The OS will have its own hostname, IP address, DNS entry etc which is separate from the physical hardware on which it operates - in some cases the OS could be operating on any number of physical components over time without any manual handling.

Are we going to start seeing things like this in desktop and laptop PCs? I suspect this will become more prevalent as people start to virtualise their desktop environments. The difficulty now becomes how to manage the OS layer separately from the physical layer and how to track this as CI data or even if its worth doing for desktops - again "it depends".

Just some thoughts.

Cheers,
Kev

[UKVIKING]

milligna,

you are dealing with 2 different things here

1 - the physical device - the PC / desktop or laptop either assign to a desk in the office or to a person who works from home or hot desking

2 - the virtual device (the instance and its details)
This device has to have an unique identification. For the finance and accoutnign and human (inhuman?) resources department - john hardesty signs for machine with the asset tag 100 - is the way that the PC device is tracked as an audited item for F&A and HR. John hardesty may get assigned a specific desk in a specific office or be classified as a float / hot desk er/

But from the IT department, this may not be the case. The PC or laptop would be given a machine name, configured for a specific O/S, have a network card - (unique MAC Address per card), get a set of Network Identification (IP address (static or DHCP), Subnet, gateway) details,
{Some details / choice may depend on what is john's work locale type).

Then comes the hard part... how is the IT deparments and its relative parts and tools going to track the PC (physical), the PC (the instance) and the user (john hardesty)

the last part is pretty easy - Active directory/domain/user
the next two depends on the tools used and how they allow for variation

for example

what happens when john hardesty gets replaced by UJ and the same machine is handed over to UJ.

Does the machine details get changed ?

IP, machine name etc ? or merely re-assigned to new person as part of a standard operational process (std change)

How does the cmdb handle john hardesty being upgrade / patched to UJ or removed completely

The config mgmt process not only needs to track the device in question but the 'owner' / primary user of the device especially for front end equipment.

As for virtual servers and devices, I am not a big fan of these
_________________
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter

[ChasingSleep]