Sergey Nivens - Fotolia
Multi-cloud storage is a great way to cut costs, ensure reliability and boost storage performance. What's not so great is when a simple management error or oversight makes the approach unreliable or unsafe.
These problems and others can be avoided with thoughtful planning and coordinating efforts within your organization. Along with overcoming internal hurdles, such as security and spending, you need to have a vetting strategy in place prior to choosing among potential cloud vendors.
Don't lose time or money by falling into a multi-cloud storage management trap. Check out these five common mistakes you should avoid.
1. Developing a fragmented cloud management strategy
One of the biggest mistakes multi-cloud adopters make is developing a plan by accident rather than by design. "When individual decision-makers choose cloud providers based on personal preference, habit or other arbitrary criteria, the organization doesn't have a multi-cloud storage strategy; it has a multi-cloud storage mess," said Rich Petersen, co-founder and president of JetStream Software, a cross-cloud data management software firm.
Rich PetersenCo-founder and president, JetStream Software
IT teams often have their own requirements and priorities and frequently adopt different clouds based on productivity needs. "The result is that IT doesn't know what data is where, who owns it and what the access controls and encryption policies are," said Anand Babu Periasamy, co-founder and CEO of MinIO, developer of the MinIO cloud storage stack. "This exposes them to a number of issues -- namely, security, application portability and duplicative storage."
There are two dangerous consequences of becoming a multi-cloud operation by accident, Petersen said. "The first is the cost of uncoordinated and unnecessarily redundant storage management operations," he said. "The second is the risk of improper data management, which could include failure to properly back up and protect some data or accidental disclosure of sensitive data."
2. Not using common data management tools across multiple clouds
Use cloud-native data and storage tools across multiple clouds, said Jeff Kim, CEO of Kmesh, a multi-cloud data orchestration SaaS platform provider. "Instead of using AWS Aurora or RDS [Relational Database Service], the customer could smartly use Cassandra or MongoDB -- open source -- which can be set up on any cloud," he said. Kubernetes, meanwhile, can be used to easily manage apps across multiple clouds.
3. Failing to analyze and understand application performance requirements
Failing to understand the performance needs of each application generally results in spending too much money on unnecessary, expensive storage or suffering unacceptably slow response.
"Analyze the storage performance profile of your cloud VMs [virtual machines] using tools specifically designed to intelligently find and report the potential bottlenecks," said James D'Arezzo, CEO of Condusiv Technologies, a company specializing in software-only storage performance technologies for virtual and physical server environments.
4. Deploying inadequate security
A multi-cloud storage security failure can be costly to an enterprise's finances and reputation. "Companies who suffer a data breach will spend millions to restore their databases to safe compliance," said John Taylor, CTO of Panzura, a multi-cloud file services provider. "Moreover, companies will also be impacted by the hurt and mistrust stemming from their customers, who feel understandably violated."
When organizations fail to create an adequate security strategy prior to deploying their multi-cloud storage management strategy, they unknowingly weaken their security posture and expose themselves to the possibility of data breaches, Taylor said.
In fact, data breaches are often far more devastating than data loss. "You can recover data, even if there is some loss, but data breaches bring with them legal liability," Periasamy said.
"IT leadership needs to acknowledge the challenges and opportunities presented by a multi-cloud approach and craft a strategy, complete with policies and controls, to ensure that they know what data exists, where, when and with what permissions," Periasamy said. "The goal is not to create barriers, but rather to ensure visibility."
5. Falling victim to vendor lock-in
Vendor lock-in is generally viewed as a strategic mistake. The damage it can cause ranges from trivial to serious. At a minimum, being locked in to a single cloud provider is inconvenient, since the customer loses price leverage, said Peter Guagenti, CMO of MemSQL, a database management software developer. "At most, they sacrifice business for their architecture, losing out on flexibility and agility or paying a competitor for cloud technology."
A true multi-cloud storage strategy enables organizations to tap into the unique capabilities of cloud-independent vendors, prioritize business over architecture and drive better costs. "But companies cannot achieve any of these benefits if they are locked in to a single vendor," Guagenti said.
The best way to avoid lock-in is to align with products and vendors that are cloud-agnostic. "Rather than choosing technology from a cloud provider, businesses must select ones that are cloud-independent, driven by a community of users and that work with all cloud providers," Guagenti said. "For example, rather than using one of Amazon's databases, businesses can choose a SQL-based database that works across multi-cloud and hybrid environments and integrates with all vendors."
Is a multi-cloud storage strategy right for you?