Over the past few weeks, I’ve personally had a sharp reminder of what life was like in IT before cloud and I thought I’d share it.
At home I run a Windows Home Server to store my media- primarily for Music but also for Videos, Pictures and Files. Windows Home Server (WHS) is a really nice bit of kit for managing the home network and for storing data it’s got a built in drive replication system so that hopefully you won’t lose anything if a disk fails. Pair that with Squeezecenter and you’ve got an audiophile quality sound network for very little money.
Everything had been going swimmingly for about 3 years, until suddenly the other day it wouldn’t boot anymore… just a very loud ticking noise when I tried to turn it on. You know, the sound of a hard drive head graunching against the platter and the “click of death”. The sort of dreaded sound that when you hear, you know there isn’t going to be any data left….
Worse, because it wouldn’t boot I thought I’d lost the un-replicated system disk, the only disk without a backup. Was I going to lose some of our digital photos, music and videos? I was in heart attack (and divorce?) territory…
Thankfully I quickly discovered that it was one of my data disks that had failed, not the system disk. Unplugging it meant the system could boot again, and I was relieved to find that all my data was there. What a relief! But because the disk that failed was 1.5tb and represented nearly a quarter of all data stored it meant the disk management service wouldn’t start, so I couldn’t add new disks or remove the failed disk.
I was totally stuck.
But a few hours lost Googling and reconfiguring Windows services, I got the disk management service to start again and then started the process of adding new disks and removing the missing one- 3 days later it finally completed.
I don’t want to be stuck in this situation again, so now the storage is healthy again, I need to do something about the OS drive, the only un-replicated drive to prevent future heart attacks. But anyone who’s dealt with storage knows this isn’t a simple task… Migrating disks, repartitioning and building a RAID array are not for the faint hearted… The RAID BIOS is characteristically unhelpful and at each stage I’m unclear as to whether I’m going to lose my data or not. I refer back to the manual and restart the process multiple times to try and make sure I’m not overwriting the disk with my data with the blank one. I’m pretty sure I haven’t as I sit here, 4 hours later waiting for the array to build with 16% done.
But more than anything what this process really reminded me of was how life used to be, before the Cloud. Endless technical issues, constant fiddling with hardware and software, and above all downtime. It’s no wonder that IT job satisfaction levels are at an all time low:
Yet another worker described bosses who expect their employees to work late into the night if need be to fix problems and then be on the job the next day at the usual time. Even vacation time is no longer sacrosanct: one person said he expects to be contacted “more than a half dozen times” during his time off.
Imagine my server was not just a home server, but an Exchange server, serving hundreds of users email. The trivial downtime issues I experienced with no access to my music and videos would have been horrendous to any organisation that relies on email. And I remember it happening to my old companies Exchange server- I still have the hard disk platter to prove it (makes a great coaster BTW).
But naturally there are fears when handing off critical business services to a third party and people are rightly hesitant, but do your due diligence right and you could end up never having the stress of fixing these types of problems again, often for an order of magnitude less cost.
The pendulum has swung and the die is cast. Cloud is the future.
