There really was only one story in 2020 – COVID-19.
During the past twelve months, the news has almost entirely focused on the major disruption that the pandemic has caused to our daily working lives and the impact that the ‘new norm’ will have on us in 2021 and beyond.
Remote management of dispersed workforces, the four-day week, the rise of the ‘digital nomad’ and the conversion of city centre offices into residential homes are all topics that are generating regular headlines. Even the tech narrative has changed from the traditional reporting of significant IT disruptions to financial services, governments etc to the impact of the failure of applications such as Zoom to the homeworker.
Yet, for all of the interest in this week’s global Worldwide Microsoft Teams Outage, the second such outage in two months, it was the release of the Uptime Institute’s third global Annual Outage Analysis report that caught the eye.
If the headline figures of the expected costs of an outage, estimated at more than $100,000 (of which almost a third of respondents reported costs of $1 million or above), with 75% of data centre operators/enterprise IT managers experiencing an IT service outage in the past three years, didn’t grab the attention, then surely the news that on-site power failures remain the biggest cause of “significant outages” will.
Why? Because unlike Covid 19, where we are in the main reacting to unfamiliar and difficult challenges every day, an on-site power failure is not a new phenomenon and is pretty much preventable.
Admittedly, the report goes on to state that power outages are declining, and that the Institute fully expects the leading cause of major disruption to be networking and software related over time, but for the time being lack of power continues to prove costly.
So, what can we do to mitigate this risk? After all we can’t control construction workers severing underground power cables, as experienced by Taiwan Semiconductor Manufacturing Co (TSMC) at their Tainan City Tech complex this week, or can we?
No, we can’t, but we can utilise the services of professional data centre operators, such as vXtream, that offer full redundant uninterruptible power supply (UPS) through onsite generators, and dual A & B 32a power feeds to server racks.
If we are collocating our equipment in a third-party data centre, we should undertake an audit of our own servers to ensure that they are configured for dual power feeds (you would be genuinely shocked at the number of organisations that run mission critical applications on single power supply servers).
And it might be an idea to actually undertake an audit of the data centre as well.
You should be seeking to determine and understand:
- the extent of separation between the A and B power feeds within the facility (ask for the building plans)
- the up-time history of the facility, going back at least 3 years.
- the frequency of power and cooling maintenance
- the length of the time the operator has been at the facility
It’s not an arduous process but a necessary one.
As night follows day, data centre failures will happen. Unplanned outages having potential serious business consequences in terms of lost revenue, lost customers, and lost brand loyalty, so the key is to become proactive in your approach to uptime rather than reactive. But that’s not the whole story. Don’t be bamboozled by glossy marketing when selecting a colo centre, expensive, expansive complex designs coupled with buzz words do not guarantee higher availability! The infrastructure requires professional and rigorous management, monitoring and maintenance.
Simply put you need an operator that cares.