How secure are virtual desktops, really?
Lock and load
Lab As we discussed in an earlier article in this series many “desktop virtualisation” solutions exist today.
Each has its own deployment architecture and comes complete with a range of operational benefits and challenges. Some work well in particular scenarios which would not suit others. With interest growing in the potential use of virtual desktops it is time to consider some of the security questions.
Little direct discussion has taken place about the security, or otherwise, of virtual desktops although there are some areas, notably more locked down environments (including public sector) and for example with respect to home working, where specific security benefits can be achieved. But, as in all areas of IT and other areas of business, security doesn’t come free or by default.
So what are the security advantages of virtual desktops versus those machines with which we have all become over familiar in the last decade or more? Whilst some are relatively clear, in truth it all depends on the particular type of virtual desktop being considered.
Clearly ‘thin client’ solutions offer all the security benefits long associated with holding all data centrally, hopefully on well managed servers. By leaving no data on the client access device, many challenges associated with desktops and laptops are made considerably simpler to address. In addition the management of the software that users employ is all held centrally and can thus be managed in a straight forward manner allowing patching and software updates to be introduced more rapidly with consequential security benefits.
There remains the need to secure the data held on the central systems, which is as always a two-edged sword – the level of risk increases with the quantity of data being held in one place, even as the risks of distributed, fragmented data storage reduce.
A raft of security benefits can be achieved with desktop virtualisation solutions that deploy out an entire VM at the request of the user and pull it back, complete with changed data files, at the close of a session. Not least that a remote computer can be lost, stolen or otherwise compromised with minimal data risk – a useful facility for both front line troops and careless business executives,
Then there are the alternative systems where a virtual machine may be resident ‘out in the field’ for some time. For these systems the problems of securing the virtual desktop are, in many ways, similar to those associated with standard desktops. Namely, data may need to be encrypted and should there be a security update for the software contained in the virtual desktop then a new copy of the updated VM must be downloaded by the user.
The ability to centrally manage the software on the desktop can help mitigate the physical challenge associated with the patching and updating software, as does the fact that the virtual desktop might consist of just a single file, or just a few at most rather than the thousands of files common in a standard desktop.
As always, and as was pointed out in an earlier article on security and virtual servers, it all really boils down to having the right processes and procedures in place to manage the systems and ensure that whatever the scenario that appropriate security is enabled. As with any other IT system, tools alone can never be the answer. Equally, making sure that users are fully aware of their responsibility and how to protect the ‘their system’ is equally important.
This is a rapidly changing area, in terms of both technologies available and best practice. So, if you consider yourself in the early adopter camp or if you have other real-world experience you would bring to bear on living with virtual desktops and securing them for production use, please do share.