Preparing for the Aftermath: Liberate Your Data!

The period we are going through right now is changing the way we work. Because IT systems have become essential to most human activities, CIOs and their teams are on the front line during this health crisis.
OpenIO
OpenIO
High-Perf Object Storage
@openio
Share
Liberate Your Data

The home office has become the norm, intensifying the use of digital tools. But it also sometimes reveals the shortcomings of business continuity plans, or the under-optimization of existing infrastructures, which have been put under a heavy load for the past several weeks. This life-size stress test obviously demonstrates the benefits of the cloud. Native cloud applications, i.e. those designed to be deployed in the cloud, have had no trouble scaling up with the addition of extra resources. And virtualization and orchestration tools make it possible to allocate these resources dynamically and configure them in just a few minutes.

Nothing will ever be the same again

Although plunged into uncertainty by this sudden paralysis of our economy, the time has perhaps come to prepare for the aftermath, as we draw the first lessons from the current episode. Let's face it: while a few companies, like Zoom, will have benefited from this crisis, most will have to cope with the need to control their spending, or even reduce it, to overcome the impending recession. Moreover, a return to normal will probably not mean a return to the previous situation. It’s obvious that after such a period of containment, getting back to the office will be appreciated by many employees. But will the office, which encourages social interaction, still be the preferred place to produce services or code?

Tomorrow, what objections can companies put forward against those who have noted increased productivity and a better work/life balance through remote working? IT professionals will have to take this digital revolution, which was imposed by a pandemic, into account.

Where to store data?

In this post-crisis reflection on IT, the issue of data storage must take an important place. Those who have not given in to the temptation of public cloud storage have noticed that it is more difficult to scale storage than compute. However, data production will intensify, resulting in exponential needs. And data exploitation is at the heart of the value creation of an ever-increasing number of activities. These include economic value, but not only: Big Data and AI have shown that they can play an important role in the fight against Covid-19, by contributing to medical research, accelerating clinical trials for possible treatments and managing the end of quarantine. Or indirectly, by optimizing the entire logistics supply chain.

However, the latest generation of object storage solutions, of which OpenIO is a part, is precisely the ideal technology for storing and serving this type of data to computing clusters.

The 3 generations of object storage

  1. The first generation was focused on archiving.
  2. The second generation specialized on the needs of sharing data.
  3. The third generation is designed for hyperscale and performance (especially in terms of achievable bandwidth), ideal for Big Data and AI workloads.

Towards an acceleration of data repatriation?

Those who had bet on Public Cloud Storage will probably have an unpleasant surprise when they receive the invoice. The cost per GB is actually quite inconsequential compared to the cost charged for the use of this data, which is difficult to predict. This oversight can lead to absurd decisions, as NASA recently migrated 247 petabytes of data to AWS... without anticipating the exorbitant cost that would result from thousands of downloads of these documents by scientists.

We can therefore bet on the acceleration of the ‘data repatriation’ movement, meaning the migration from public cloud services to private cloud infrastructures. These can be on-premise or hosted by a cloud provider.

By repatriating their data, companies benefit from the best price/performance ratio, and retain complete control. The main obstacle that companies face is mobilizing the skills required to deploy and manage a storage cluster, either internally or via a provider. That is why OpenIO is offering more and more managed services. One example is our support service called "You Host, We Manage". This is a storage management service dedicated to companies with large clusters of at least 5 PB, where OpenIO manages the cluster so that our customers can focus on their core business.

The end of corporate NAS in favor of collaboration platforms

More and more often, we see Object Storage replace NAS servers that were previously used to deliver data to web applications. Replacing NAS with object storage allows companies to lower their costs while increasing their infrastructure resilience (see the Dailymotion use case).

The explosion of remote working has accelerated the obsolescence of enterprise NAS servers (NAS scale-out), which have long been used to share documents between members of an organization connected within a single private network. Remote and mobile communications require new collaborative tools, such as Nextcloud or Storage Made Easy, which allow people to access, share and edit resources from any device (computer, smartphone, tablet...) in a collaborative way. Object Storage can be used as a low-cost and scalable back-end storage technology for collaborative platforms. In comparison, enterprise NAS, which is often based on appliances and file systems, results in a high TCO and limited capacities.

Infrastructure modernization and fleet optimization

Crises are a breeding ground for innovation. The current situation will have shown the role technology plays in maintaining the link between individuals. The situation will also lead companies to modernize their infrastructures, as the crisis will have accelerated the adoption of new tools that will most likely become a lasting part of working habits.

The way we propose Object Storage, defined as a hardware-agnostic software, could prove to be an intelligent solution to reuse older generation machines, and increase storage capacities at low cost. Especially since OpenIO also perfectly supports hardware heterogeneity within the same cluster.

The future belongs to those who will make the most relevant choices before the others.

One question remains: in times of crisis and uncertainty, is it reasonable to launch a migration project to a new storage solution? After the rush of the first few weeks, your teams are probably less stressed and more available. Above all, this crisis, by taking everyone back to the starting line, is also putting the cards back on the table. The future belongs to those who will make the most relevant choices before the others.

OpenIO
OpenIO
High-Perf Object Storage
@openio
Share
All posts by OpenIO