Downtime! How to avoid it
In the last year, many businesses have moved to a managed service provider (“MSP”) like HostSailor. They did so for a variety of reasons, mainly for cost. Retail, in particular, has been badly hit with shops closing and income plummeting. Their only survival option is to move to an online presence with an e-commerce platform. Does downtime have a role to play in e-commerce or any other area? In many cases, cost considerations and a need to implement quickly mean that an MSP is the only option. All businesses with internal IT systems experience it for a variety of reasons. It’s a fact of life that the infrastructure will fail from time to time, usually at the worst possible time. MSP’s, while trying not to, do also have it.
Downtime Types and Reasons
The first thing to understand is that downtime is sometimes outside the MSP’s control. Not being able to access systems through an Internet connection does not necessarily mean that the systems are down, just that your connection to the MSP has failed. Internet connections pass through a series of third parties, and if any link in the chain is broken, the entire chain is broken.
What also needs to be understood is that there are several classes of it:
1- Catastrophic.
The entire Data Centre is inoperable, usually because of a power outage. There should be backup power arrangements, but these may only last for a limited time.
2- Total.
The Data centre is up and running, and you can connect to it, but your business systems are wholly unavailable or unresponsive;
3- Partial.
The Data centre is up and running, and you can connect to it, butparts of your business systems, perhaps integration with other systems, are unavailable or unresponsive;
4- Irritating.
You can carry on with normal business operations, but some functions are not available.
Common reasons for downtime include:
1- Equipment failure.
Something as simple as a cooling fan in a piece of equipment failing can shut it down.
2- Software failures.
Some software systems automatically download and install upgrades. From time to time, they can cause the systems themselves or their integrations to fail. In very rare cases, operating systems close down.
3- Malicious activities.
Targeted attacks on the Data Centre and malware can cause downtime.
4- Force Majeure.
In some countries, government restrictions can cause Data Centres to close entirely or restrict their connectivity.
How the feeling is during the downtime! A banner warns against this points refers to “Avoidance”
For example, in Ethiopia Internet connectivity is supplied through a single source, the parastatal, Ethiotel. During civil unrest, the Government has instructed Ethiotel to, simply put, switch off the Internet and cellphone services.
Avoidance Measures
1- Power loss.
The Data Centre has backup power arrangements. These are usually at two levels, batteries to cover for short outages, and generators for longer-term power losses. Backup planning must include a provision to refill the diesel tanks, and what to do if a generator fails.
2- Equipment Failure.
This is usually following an equipment failure in a critical device. It can be as simple as a cooling fan failure. Most key equipment has hot-swappable components and failure alerts, minimizing potential downtime. Using Raid can protect disk storage against a single disk unit failure. SSD units have a limited lifetime but are immune from mechanical failure.
3- Malware.
The FBI has recorded an upsurge in malicious activities over the last year. DDoS attacks have increased in number and viciousness, and they theorize that could be due to online merchants trying to block competitors.
Other vectors like Phishing have shown similar increases, and fraudsters try to take advantage of the confusion. An MSP must have comprehensive antimalware defences in place and a shared environment insulate one customer from the actions of another. Network monitoring will show up network attacks to be blocked before they cause a loss of service.
4- Force Majeure.
There is little you can do unless there is an alternative route to get around the issue. One technique is to have a second Data Centre in another jurisdiction as a hot-standby and route all customer traffic to that site. MSPs like HostSailor with sites in multiple countries are better positioned to offset local actions.
MSPs like Host Sailor take uptime and service levels very seriously. When downtime happens, it reflects not only on the customer but also on the MSP, causing great reputational damage. They make great efforts to provide a safe and secure environment for their customers.