Redundancy in the case of data center infrastructure refers to the act of duplicating a component to ensure the seamless operation of IT equipment even in case of failure of a component. A good example of this is having a redundant power backup to ensure the continued functioning of the data center in case of a power outage. Redundancy is mostly practised to make sure that there are zero downtimes in a data center, even during the worst of situations.
The situation pre-Covid
The situation for hyperscale data-center operators was moving towards an exponential rise in demand for redundancy long before the pandemic hit. The redundant physical infrastructures in data centers also meant that the IT processes and networks were also getting redundant in order to ensure continuous power supplies.
Operators were already rethinking their approach towards data redundancy prior to the pandemic and steps were being taken towards getting the costs of redundancy under control.
Before the pandemic, cost reduction was focused on building facilities that had less equipment redundancies and had “multisite availability zones” within the network. Multisite Availability can be simply described as the process of shifting the focus from infrastructure to IT systems in order to create redundancy.
The impact of the pandemic
The impact of the pandemic on hyperscale operators and cloud providers has been to raise the stakes and to challenge the traditional approach to reliability. Multisite availability zones were showing vulnerabilities like network and software failures even before the pandemic but the pandemic made these vulnerabilities more prominent than ever before.
With the pandemic, creeping criticality in the IT systems has become higher than ever before. Creeping criticality can be described as a situation where the infrastructure and processes are not updated to the level of the criticality of the applications that they are dealing with. This has also meant a greater number of IT system failures than ever before.
As a result, hyperscale and cloud operators have moved towards managing concurrently maintainable facilities and demand N+2 infrastructure from colocation partners. With the pandemic, as the growing work-from-home demands and remote data-center management exposed the weakness of virtual-redundancy set-ups, the move towards concurrently maintainable facilities was greater than ever before.
What to expect in the future
The future of the data centers demands that greater focus be directed towards increasing the physical redundancies of a data center.
A key factor towards achieving this is by introducing smart components such as rack PDUs or rack-mount PDUs in redundant configurations. This setup should be included in situations of concurrent maintainability reaching rack-level redundancy.
Smart and simple PDUs when used in an intelligent fashion can play a very important role in supporting critical power needs and making the data center more prepared for the post-pandemic world! If you too want to be better prepared, simply reach out to us here and we will be happy to help!