IT Asset Management has taken up quite a bit of my time over the years. Creating the “IT Inventory” is often a challenge in an organisation, building and keeping accurate records of hardware and software once the number of devices grows can be difficult and isn’t a “do-once and forget” task. It’s common to start with a discovery session – either by scanning the company network or sending someone round with a clipboard. For me, however the first step is all about the questions; what do we want to ask the database to tell us?
The need for an asset database, inventory, or CMDB can be driven from many different directions. At work I’m currently involved in a project being pushed very much from the senior management level- the board want to know the answers to things like “how many computers do we own?”, “how much did they cost?”, “are we properly licensed?” and “are we disposing of old computers properly?”
From an IT Support point of view the questions are different. Analysts want to know the answers to questions like “how many Windows XP machines should I get round to upgrading before April 8th, 2014?”, “How much memory is in PC1324?”, or “Joe Bloggs has just called, what is the hostname of his laptop?”
The questions are important. The questions determine what information we need to collect both in the initial discovery session, but also (and in many cases, more importantly)over the years that the inventory service will run.
For example in the questions above we can answer “how many computers do we own?”, and “how much did they cost” by only recording a serial number and a price. However, if you also want to know “How much memory is in PC1324” then your service will need to record the RAM in the box. “Joe Bloggs has just called, what is the hostname of his laptop?” would also need us to record the “owner” of each hardware asset as well.
There are many questions that can be asked, and I wouldn’t expect the list to ever be comprehensive, but it gives valuable insight into what the users of the service want it for. Whilst it is possible to design many services to be extensible and to start collecting additional information at a future point (for example we can start interrogating hosts on the network to find out how much memory they currently have installed) unfortunately some information can be lost for ever, or at least difficult to import (if the original cost of a PC was never linked to it’s asset tag we may never know how much PC1324 cost to purchase).
My advice would be to find the Stakeholders and then find out what questions they want to know the answers to. From there you can work out what data needs to be recorded, then it’s just the simple case of working out the how, where, and when!