The post How does point in time recovery work? appeared first on Over the Wire.

“A Service Level Agreement is not a guarantee that your infrastructure will remain active for the specified uptime. Rather it is the minimum period of outage after which the business will receive a financial rebate, where this has been stated in the contract.”

For instance, a 99.95% uptime SLA (which is typical for the majority of business grade Ethernet services) has a maximum target downtime of 21 minutes and 55 seconds per month. Many IT Managers make the costly mistake of thinking this means that the service is guaranteed to be active for the remaining time period. It isn’t. The SLA simply means that come 21 minutes and 56 seconds, the customer will be able to claim some remuneration for any additional downtime experienced.

Why are the SLA’s not guarantees?

Using the example of a 99.95% SLA, 21 minutes and 55 seconds is more than enough time for a provider to fix the majority of network faults. However, infrastructure faults, which often rely on third parties to be resolved, can take much longer. For instance fibre cuts, ULL faults, and NTU failures are normally beyond the scope of the service providers control and as such require third party technicians to be physically present at the point of failure to be resolved. Depending on the fault, the remoteness of the area, number of services affected and the availability of qualified technicians, infrastructure failures may take several days to resolve.

Why does this matter if the provider has to compensate me?

Whilst an SLA may provide remuneration in the event of a service outage, the rate of remuneration is determined by the provider and is not linked to the actual loss incurred by the business. This means that the risk of a failure still needs to be weighed against the potential costs of one occurring. If the impact on the business is deemed to be unacceptably high it is strongly recommended that organisations consider utilising a redundancy package to mitigate the risk of an outage occurring.

What redundancy options are available?

Typical considerations for redundancy should include the following; however, it is recommend that all organisations consider backup links on a per-site basis so that the appropriate technology can be balanced against the cost.

• Due to their low cost and geographic availability, ADSL links are often recommended for the majority of sites.
• However if the site is in a remote location or there is a high risk of a large scale problem (e.g. flooding, bush fires, etc.), 3G and 4G backup links can be used. 3/4G is also recommended as a method of providing redundancy against terrestrial problems. For instance if a fibre cable, or other form of cable infrastructure is cut, it is likely that an ADSL connection will be affected as well, making 3/4G the most viable alternative. Microwave transmission can also be used to the same effect, however it comes with its own advantages and disadvantages. You can read more about the differences in connectivity options in this article on connectivity.
• The deployment of a second router to provide additional device redundancy can protect against hardware failure and get you back to work almost instantly.
• Due to a slow reduction in costs, protected fibre (fibre that takes diverse routes to the exchange) is becoming increasingly common as a backup option. Protected fibre comes with the advantage of being able to offer the same speed as the primary connectivity option.

Considerations should also be made as to whether there is a requirement for automatic or manual failover, as each of these options has the potential to affect the speed at which your service can be restored.

The next time your organisation is considering setting up a link to a new office, or upgrading an existing network, make sure you keep in mind that an SLA is not a silver bullet, and that the risk of an outage needs to be balanced against the potential costs of one occurring. 

The post Don’t Get Caught Out By Your SLA’s appeared first on Over the Wire.