ICT Management. > Strategy & Planning
Software > Buying & Owning Software, Internet Based Software

Moving to the cloud - a case study

By Morgan Killick

< Previous | 1 | 2

Decommissioning Existing Infrastructure 

A further hurdle was the business of decommissioning the server, which ESP was more fully involved in. If the server was to be eliminated, they would need file storage, and their PCs must leave the domain. They knew it was not wise to put 1000’s of accumulated documents – many of them containing personal information - onto Google Apps. But equally with 12 users, a consumer-market NAS was not appropriate. We suggested instead a more suitable RAID NAS with its own UPS for power protection. This was £600 in itself.

Leaving the domain was going to be quite a task too. They were happy with Microsoft Office, and having already paid for it, saw no advantage (or cost savings) in using unfamiliar online office suites. This meant that ESP would need to reinstall and reconfigure Office locally as it had been centralised years previously and was currently deployed by Group Policies. More than that, we also needed to reinstall or redeploy printers, other common applications and the entire antivirus system on all 12 PCs, as well as reconfiguring the backup. The migration also meant a new permissions structure, creating new local profiles on every machine and re-importing Desktops, MyDocuments, Outlook profiles, Favourites and other settings - another huge tranche of work. How would the roaming profiles work? Answer: they wouldn’t! Sure, they could log on to Google from any PC, but their permissions on the NAS box wouldn’t follow them around and nor would their MyDocuments and other Desktop files, icons and folders. Password enforcement on PCs was now out of the question so their computers could end up with all sorts of different user accounts on them to allow volunteers and others to ‘hotdesk’.  

All of this meant substantial disruption. When you start to look at the detail of reverse-engineering an entire Windows domain, you realise it’s a big job – we estimated up to 2 days of work with an awful lot of disruption to all core systems.

A Step Too Far!

Following these protracted discussions and with the migration date drawing ever nearer, we received an email from the client saying that they had decided to postpone the work indefinitely. Their reasoning was as follows:

• Cost-wise, the project as it stood had failed to pay for itself, even over a few years -  The revenue savings on the ‘server element’ of their support contract for instance were not actually that great and on top they were facing considerable upfront costs too what with the significant amount of labour required to make the system work without the server, plus the cost of the NAS device.
• The sheer amount of work for them and the disruption involved in the project, was too much to bear, especially at a time of organisational uncertainty. There were so many changes to go through and too much time and effort from senior staff was needed.
• They knew the testing phase was incomplete at best and they had never really overcome the fear that something vital would be lost. Frankly, they were worried that the new system wouldn’t do the job and too much faith had been vested in key individuals who weren’t really sure they knew what they were doing.

On top of this I would add that the impact of losing pretty much every network management feature had been underrated. It became clear to us at least that disposing of this function altogether would mean a return to an unmanaged network with no control over users or computers. The cloud – it seemed – had passed. For now at least.

Conclusion

Cloud computing certainly does offer many benefits to all sizes of organisation. We have found though that ‘middling size’ organisations with one or two servers in particular have a lot to think about before attempting wholesale changes, and indeed whether it is worth even trying to do this at all. There is a mantra in some circles that ‘moving to the cloud equals cost savings’ and this may well be true over the long term. Yet there are unspoken barriers here – chiefly significant immediate expenses and disruption to systems - that third sector organisations need to be aware of. Moreover this equation barely applies at all in the case of organisations that can apply for funding to replace servers, or if existing  infrastructure can be eked out for longer by making smaller, more organic changes.

We should not forget that the concept of ‘dumping servers’ and firing the IT Support staff originated in very large companies with sophisticated IT infrastructure. For those of us with more humble technology in the office, ‘reverse engineering’ an entire Windows domain and moving all of those ‘lost’ system functions onto the cloud - or living without them - is a big undertaking. It will invariably involve significant input from you and your team to learn, plan, test, teach and adapt.

A more realistic approach is to start gradually, appraising one technology at a time and remembering all the while that cloud based services continue to mature. Cloud based office suites (Google Docs, Zoho Docs, Office 365 etc.) and data storage / backup/replication solutions make sense for a huge range of organisations but they won’t remove the need to retain some functionality, software, hardware and support in-house.

Finally, unless you happen to be very IT savvy, you will still need IT consultants! Cloud computing, like any technology, needs experienced, reliable expertise and this will come at added expense. And if those consultants are to inspire your trust and confidence, it is vital they that they have extensive experience of both cloud solutions AND client/server networks too!