Throwing The Availability Lifeline

Andy Bailey

Andy Bailey, 12/03/2010, posted in "Analysis"

Andy Bailey is Availability Architect at Stratus Technologies. When not blogging about High Availability, Continuous Availability and Fault Tolerance, he enjoys fast cars and relaxing with his Pipe Organ. ...more info

Andy Bailey is Availability Architect at Stratus Technologies. When not blogging about High Availability, Continuous Availability and Fault Tolerance, he enjoys fast cars and relaxing with his Pipe Organ. ...less info

• Tweet
• Email

• Tweet
• Email

Ummm…..so what did you think of spending an hour in the middle of the working morning to tune into The Register’s web cast I promoted earlier in the week? From the start I was confused…

The event had been billed to include virtualisation, security and availability. As it went live we were invited to download a copy of the slides “High Availability and Disaster Recovery”. Upon downloading, the slides were entitled “Virtualisation and Security”. What was going on?

Anyway, I decided to stick with it, started to enjoy my pre-prepared coffee, turned the phone off and tried to ignore the usual flood of emails and tweets, hoping that the vital role of availability in the virtualisation journey would be revealed.

Now don’t get me wrong, there were a couple of useful pointers, for instance:

• The realisation that a virtual machine was so much easier to steal than trying to physically walk through the corporate turnstile with an hp380 (let alone a Stratus system) under your arm.
• The realisation of how easy it is to ‘run up’ a virtual machine for a couple of minutes for a quick test and inadvertently infect the rest of the estate. By the time someone from IT security pinpoints the source the VM will have been long since switched off (I don’t envy the security experts having to track such occurrences down).

But those slides – wordy, wordy, wordy!

And what about availability? Absolutely zilch. Nothing. Very disappointing.

I felt like I’d tuned into the wrong event so didn’t feel obliged to ask any questions. On reflection, perhaps I should. Perhaps I should have thrown Tim, Neil and Tony a lifeline, asking:

• Can you describe some examples of best practice of how to achieve high availability please?
• How can I ensure that services are not disrupted in the first place (don’t forget good people, all the HA stuff means you have downtime and have to restart services)?

Of course, working for Stratus I have plenty of examples of the first and know the answer to the second is “build it on a fault tolerant platform”. I know these guys know this too. Shame they overlooked it…

Possibly related posts: (automatically generated)

• New Year Prediction? It’s going to be cloudy with a chance of FT…
• High Availability in the Decade to Come
• Market News December 15 2009 Market