Before you leave, get your free copy of our Data Center Migration Checklist
Use this checklist to help protect you investment, mitigate potential risk and minimize downtime during your data center migration.
Imagine that you just found out your data center is moving from its metro location near your office to the suburbs. At first, you may wonder why they would do this considering their clients are nearby.
After all, enterprise data centers from the early 2000s were typically built in metropolitan areas so they would be closer to their customer’s corporate offices. During this time, power and real estate were less expensive and it made sense.However, as costs increase, some data centers are considering moving. As data centers transition from urban locations that are close to their customers, to newer facilities to cut costs, it can create new challenges.
In order to support customers needing connectivity associated with on-demand cloud consumption models, suburban data centers may need to increase their inter-metro bandwidth. This can require an upgrade to faster 10 and 100 GbE services.
Since most data centers are packet-based and use Ethernet, they can benefit from using hardware that supports very dense 10 GbE that allows them to maximize bandwidth with the least amount of power within a confined space.
Most data centers are designed with the assumption that equipment is located within the building (or in a nearby building). This allows for easier troubleshooting since all devices are within easy hands-on access. However, once traffic goes outside the building into an optical fiber network, it can run underground or on telephone poles which can be far more complex to troubleshoot. This can increase the risk of downtime and should be mitigated using multiple carriers.
Metro networks use high-capacity coherent optics and photonics. This allows for faster transfer of data. “This availability requires a rich set of optical and packet operations, administration, and maintenance (OAM) tools that allow operators to proactively and reactively troubleshoot their network”, says Mitch Simcoe with networkcomputing.com. This allows problems to be fixed quickly so that data centers can maintain service-level agreements with their customers.
Using the converged approach allows data centers to combine “optical and packet layers into a single network platform and locating IP routers only where needed, such as within the data center”. Using this method can be more cost-effective because it allows data centers to switch slower traffic within the metro network. “We can achieve massive scale with this approach, while maintaining the benefits of Ethernet – from both a simplicity and cost perspective”, says Mr. Simcoe.
Since Ethernet has become a dominant technology within the data center, it makes sense to use networks that are able to rapidly switch and aggregate Ethernet. This will help reduce complexity and cost when deploying “high-touch functions” that were not be needed in their previous metro location.
As metro data centers move to more suburban areas, there will likely be challenges; additional planning will be required for bandwidth and troubleshooting may take more time as it requires tracking down issues outside of the data center. Using a converged approach, data centers can allow cost-effective switching of slower traffic within the metro network, which will allow them to better scale in their new location.