So it’s the first day of Spring, the sun is out, the hail is out, it’s snowing – no-one really knows what’s going on with the weather.
This unpredictability is because there is nothing out there that can 100% accurately say what’s going to happen today. This, luckily for you – is NOT the same case for your servers.
Servers are arguably going out of fashion quickly due to cloud computing (a whole other barrel of fish, we wont get into it now) – but for those that still have them whether due application support, recent purchase of Exchange or whatever it may be – monitoring is the be all and end all of an IT Administrators job. It’s the big, important, noisy box that your business undoubtedly depends on.
We often take over a company’s IT and the first thing we look at is the server. Often, it’s a complete mess. Random stuff installed (do you really need the Ask toolbar installed? I mean not even on your PC, definitely not on the server!)
At Vital we have a very standardised set up of our servers, complete with checklists for roles installed, shared folder locations, firewalls etc – this ensures that when we use monitoring via our Remote Monitoring Platform, everything is the same so our techs know where things are and if something does go wrong, can fix it swiftly and easily based on our system knowledgebases.
Now this leads me onto our standards… virtualisation, perhaps being one of the biggest items we consider and something that has saved us numerous times. For those that aren’t aware, virtualisation essentially allows you to run multiple servers off a single physical box – rather than having a different physical box for every server (which is very space consuming, noisy and just very unnecessary these days). With using primarily Windows systems, we use the same on our physical boxes – a Windows Server OS running Microsoft’s Hyper-V product.
We literally install the bare minimum on the physical – the latest updates, the roles Hyper-V, SNMP and Telnet. SNMP is very important as it allows us to monitor the hardware of the physical, rather than just the usual OS related items like disk space, updates, CPU, services etc – SNMP let’s us tell you when your server is too hot (something a lot of our clients like when we ring up in the summer!), when a disk fails or even when one of the power supplies fails etc.
However the reason we install the bare minimum is because in the worst case scenario and that physical box OS fails, we don’t have to worry about any data being lost – we just reinstall the OS and install the roles again which, if we don’t do updates instantly, can take us around 30 minutes. This means we can have that client up and running in the shortest amount of time possible and the smallest amount of disruption.
Essentially a virtual machine is a single file (or as many separate disks as you want). Rather than having a physical hardrive, you have a “VHD” or “Virtual Hardrive”. This file contains all the information from boot partition to OS to data etc. In order to make it easy if the OS fails, we separate our OS installation disk and our data disk – again, just another added layer of protection if something were to go wrong. It does mean however, if the physical fails completely or you upgrade to another server etc – these files can simply be moved over the network or via USB etc and then we do something not too fancy, connect it to the Virtual Machine and voila – it’s back up and running.
Now this takes us to backups – everything we do is just an added layer of protection – rather than having all our eggs in one basket, we have multiple recovery solutions depending on what fails, if it fails or if it could fail. Backups are one of those things that most people don’t worry too much about unless they have already had a catastrophic failure and THEN realise how important those backups are.
In fact, below are a couple of backup statistics that you may be blown back by:
- According to a survey by the Strategic Research Institute, companies that are unable to resume processes within 10 days of a natural disaster are unlikely to survive more than a year. According to another survey by Baroudi Bloor International, half of all companies that do lose their data go out of business overnight.
- Of those businesses that do back up their data, 34% never test their back-ups and of those that do 77% are found to not work. Backup is only half the solution, make sure you can recover.
I wont go into backups too much (potential idea for my next blog!), but virtualisation is NOT a replacement for a backup solution. What it is, is a very handy tool that can reduce the amount of work for an IT Administrator and most importantly, allows the administrator a number of added benefits for general management; a few below:
- During a server restart, we can monitor the progress through BIOS to OS boot by connecting via the Hyper-V console – this means that if the OS fails to boot – we can find out why without having to travel on-site first – this eliminates a good amount of time and means we can diagnose and fix the problem much quicker.
- Because a Virtual Machine is just that, virtual, it means hardware support is much easier. When we provision a VM, we give it the minimum amount of hardware required – this means if there are performance issues or something needs changing, we can increase the memory, increase the disk space, add a network card or whatever it may be much quicker – again without us having to be on-site and in some cases, without even having to take the server offline!
If you need any assistance or think you may need a “fresh look” on your server administration, or maybe your current IT company isn’t proactively monitoring your system – then give Vital a call or complete the call back request below and we'll call you. We’ll make sure to have the best solution in place for you, whatever IT takes!