• Spark has launched dedicated Auckland data-centre links at 10G, 100G and 400G, with 800G described as a near-term addition
  • The service uses fixed optical paths without restoration, so customers needing automatic recovery must plan for another route or transport service

The fact

Spark Wholesale launched a new data-centre interconnect service in Auckland on 8 September, connecting facilities between Silverdale and Papakura. Customers can order dedicated links at 10G, 100G and 400G using Ciena optical equipment.

The company also said that 800G service will be added soon. Its product page already lists 800G as an option, so customers will need to check whether it is available for the locations they want to connect. The underlying Ciena equipment can carry up to 1.6Tbps, but that is the capacity of the optical technology, not a 1.6Tbps service currently being sold to customers. The network also uses Ciena software to help Spark monitor capacity, provision services and troubleshoot faults. For now, the new interconnect service is focused on Auckland rather than being available nationwide.

There is also an important difference between this service and Spark's wider Data Transport Service. The new data-centre links use fixed optical paths and do not automatically switch traffic to another optical route if a path fails. Spark says multiple fixed paths are available at some locations, while its separate Data Transport Service includes restoration over available alternative paths and is offered nationwide.

The assessment

Customers linking two Auckland data centres need to think about more than speed. They also need to know what happens if the connection fails. Spark’s DCI service uses a fixed optical path, so traffic will not automatically move to another route if that path goes down.

For some customers, that may be fine because they already have a second connection or manage failover themselves. Others may want Spark to provide that protection. The important point is to check whether a separate route is available between the two sites and who is responsible for moving traffic during a failure. A faster link does not solve that problem by itself.

For BTW readers, the choice should come down to both capacity and resilience. A 400G or future 800G fixed link may work well where backup connectivity is already in place. Customers that depend on the carrier for recovery may need a different service. Spark has not published pricing or detailed recovery terms for each location, so those details will matter when comparing the options.

What to watch

Watch for confirmed 800G availability and details of which Auckland facilities can be connected through separate physical paths. For customers using the service for replication or disaster recovery, written route-diversity and restoration terms will matter more than another increase in maximum bandwidth.