If you thought identifying devices like PCs, tablets and smartphones on a network is just another basic task of a modern IT department, think again. In fact, effectively managing fleets of devices requires considerable effort and reliable tools. That’s exactly why device management software like Miradore is such a lifesaver for IT managers.
One of the key tasks of IT asset management is to keep you up to date on the devices your company owns. But can you really trust your asset management system to succeed in this, as the devices move through their often very complicated lifecycle? Let’s take a look at some of the possible pitfalls.
The urgent need for new identifiers
In a networked environment, for small companies and enterprises alike, the most readily available and widely used identifier is probably the device hostname. This works perfectly until, say, your employee gets a new laptop and the old one is stored away for a while. The trouble starts when, in a few months, the old laptop is given to a summer trainee and it’s reinstalled with a new device name. How many devices was it you owned?
A BYOD program brings an entirely new set of challenges. If users set up their phones like the manufacturer intended for home use, you will have a fleet of Peter’s iPhones and Mary’s iPads instead of a consistent naming policy. Several devices will share the same name, and with some manufacturers the names might even be in different languages. We clearly can’t rely on the old identifiers for company devices.
The trouble with hardware attributes
In a perfect world, device serial numbers and the MAC addresses of network adapters are worthy candidates for the gold standard of device identifiers. They are mostly based on hardware and can be easily read using software.
If only these attributes were always unique, but alas, hardware manufacturers have different practices. While attributes based on hardware are more stationary that device names, they might still change, for example, due to the replacement of a broken network adapter.
Miradore comes to the rescue
Thankfully, starting with Miradore version 4.1.1, our customers can breathe a sigh of relief. While our software used to rely heavily on device names, we have now put our comprehensive hardware profiling to good use. Miradore now looks for a combination of the device name, serial number, MAC addresses and DNS domain whenever a newly installed device greets the server for the first time.
At least three of the above attributes must match for a device before Miradore will consider it an existing asset. There are some minor exceptions – the limit is lowered to two, for example, if Miradore hasn’t received the device’s DNS domain. If the Miradore database only contains a device name and domain, a device is only considered a match if both these attributes are identical.
After a new device has been registered to the server, it gets a globally unique identifier. This is called its client GUID, which will identify it in the future. It will also allow Miradore to keep track of identifier changes, even if multiple identifiers are changed at once and the device can’t be identified on physical attributes alone. In the event of a replaced motherboard, for instance, Miradore can recognize the device regardless of changes to its serial number and MAC address.
The third party challenge
Yet another challenge with identification concerns third-party systems, like Microsoft Active Directory or antivirus management servers. Miradore can talk to them using connectors, but these systems tend to provide limited sets of attributes for identification.
For example, the Active Directory connector currently identifies devices by their name and domain. Other connectors use only the device name. Fortunately, these systems are typically used in environments, which use consistent naming policies, so there is little chance of devices sharing the same name.
Our work will go on and on
Automatic identification is a crucial timesaver for IT managers, but no system is infallible. We accept there is always a small risk of identifying devices incorrectly, which is why we offer and build tools for resolving such issues manually.
We see device identification as a critical part of making IT environments more consistent and accessible for our customers. Consequently, we are continuing our work to smooth out the remaining challenges, and we’ll keep you posted with any breakthroughs we make.