How to deal with the challenges of cloud storage vs. SAN
Whether you're managing cloud storage or a local SAN, both have issues. Compare 13 areas where SAN and cloud storage challenges can arise and find out how to deal with them.
By moving data storage to the cloud, businesses can focus more on the core competencies of supporting their business model.
For instance, cloud services simplify the data center backup and recovery processes, enabling a smaller staff to manage them. You no longer must manage a large backup infrastructure but, instead, can handle recovery with a few mouse clicks. Large cloud providers offer data redundancy across geographically disparate data centers, in turn, making your data safer.
Among the greatest benefits of the cloud are the rapid deployment of resources and timely decision-making. These advantages are particularly relevant today with COVID-19 and so many people working from home. Entire environments can be spun up in hours versus months. Cloud services are perfect for rapidly changing workloads and quick responses to emergencies.
Nevertheless, there are many challenges when extending the storage infrastructure into the cloud. What follows is a look at 13 areas where challenges arise in managing cloud storage vs. a local SAN, including:
- Cost management
- Migration plans
- Cloud data repatriation
- Service-level agreements
- Customization
- Security and compliance
- Architecture
- Compatibility
- Backup and data recovery
- Support
- Vendor lock-in
- Manageability
- Reliability
Key local SAN and cloud storage challenges
Cloud storage challenges often relate to the choice of a cloud provider. With a local SAN, there's more variability and flexibility in dealing with problems. It's important to know how to handle the challenges of cloud storage vs. a local SAN. The best approaches require due diligence and a proactive process to make the correct choices for your business.
1. Cost management
Costs are difficult to plan for and manage with cloud deployments. The ability to rapidly expand resources is a key cloud advantage in the cloud storage vs. SAN debate, but costs can skyrocket when it's so easy to create new environments.
Do an audit to estimate your cloud infrastructure costs, including the ones involved with:
- executing the move to the cloud;
- moving any data back on premises from the cloud; and
- additional post-migration costs.
Outbound data transfer is charged at the normal rate. Inbound data transfer is free. It will cost more to bring servers and data back on premises than to move them to the cloud. There's no direct comparison among providers because each offers different options, including pay as you go, reserved volume discounts and instances per second used.
Obtaining automated and detailed usage reports for cost tracking and reporting is challenging. Third-party management and reporting tools, such as Datadog, Dynatrace and SolarWinds, can be helpful.
Cost management in a local SAN environment centers on hardware lifecycle management, service contract management and managing the capital expenses that go with those items in the budget. Lifecycle management and proper planning of hardware replacements is key to reducing costs when managing a local environment; hardware support costs can spiral after five years.
2. Migration plans
Other cloud storage challenges come up when considering migration issues. A cloud migration plan must consider all the workloads to be transferred to the cloud and the sequence for migrating them. Consider using professional services to help design and build a step-by-step approach, which, in turn, enables the implementation team to learn as it goes and to create detailed documentation.
A local SAN migration plan must consider the underlying network infrastructure and compatibility. For instance, the old SAN or Ethernet network might be 10 GB and the new 25 GB. Software migration tools can help. Vendors often provide these tools free, but they can be difficult to use. Third-party tools, such as Datadobi, also are available. Migrating to an entirely different hardware platform requires attention to security, protocols, IP and name changes, and shares. Proper planning is key whether planning a migration in the cloud or locally.
3. Cloud data repatriation
Why move back from the cloud? Many factors may drive the need for some resources to move back on premises. Keeping data in the cloud can become less cost-effective over time. Other issues can come up related to data control and security, performance, I/O requirements and vendor lock-in.
Proper planning for high-performance applications or ones that have specific data-compliance requirements is critical. Some of these applications are better suited to the local SAN environment.
4. Service-level agreements
There's no standardization of cloud service-level agreements (SLAs). Each provider has its own SLA metrics, restrictions and exceptions related to service availability and standards. Consider your organization's availability, response time, capacity and support requirements. Pay close attention to all legal requirements for data security and determine who's accountable for failures.
Local SAN management requires custom SLAs based on local resources and mutual agreements between the business groups and the SAN management team. It's important to review and create manageable and realistic agreements that the local IT team can support with back up from vendors.
5. Customization
In a cloud environment, security risks are important to consider when customizing for the integration of mobile deployments, deployment of content in multiple markets and the handling of personally identifiable information. Legal agreements may be needed with a cloud provider to ensure the compliance of any customization.
The same considerations apply with a local SAN. Proper planning for business applications is critical, and the environment must be customized for specific application needs, such as disk I/O and network bandwidth. However, it's up to the local team to ensure the environment is properly customized for all business requirements.
6. Security and compliance
Other cloud storage challenges relate to security and compliance. For one thing, cloud computing widens the potential attack surface. While the cloud provider is responsible for physical security, business continuity, DR and network security, additional security controls and responsibilities are left to the customer.
Security and regulatory requirements often dictate what can and can't be moved to the cloud. With the numerous state, federal and international requirements -- including Sarbanes-Oxley Act, Criminal Justice Information Services, HIPAA, GPDR, and the Family Education Rights and Privacy Act -- it may make sense to store data locally.
When considering cloud storage vs. SAN, security requirements are similar. Again, it's the difference between a vendor contract ensuring compliance and a local team making sure the requirements are met. You need to understand your security goals, what each provider offers and who's responsible for what.
7. Architecture
From a cloud deployment perspective, consider your environment and how the provider's offerings fit in to your workflows. Microsoft Azure may make sense for an organization that's already using a lot of Microsoft software. Amazon or Google cloud services might make sense if you use either vendor's other services.
From a local SAN management perspective, architecture design hinges on striking a balance between price and performance to meet business needs. Other considerations include business continuity, security, reporting, backup and recovery.
8. Compatibility
When looking at the challenges of cloud storage, compatibility is more of a cloud deployment issue than a local one. In the cloud there are more SaaS and PaaS options. Application compatibility testing is critical, but it's also complex. When testing identifies compatibility issues, other options such as infrastructure, PaaS and SaaS can be considered. Ensure that an assessment and compatibility test is completed prior to any migrations.
9. Backup and data recovery
Business continuity and data recovery can be easier in the cloud because of built in recovery options. The biggest challenges for data recovery from the cloud are related to operational recoveries, including bandwidth consumption, the cost of bringing data back and the time it takes to recover data. Many critical workloads may need to remain on premises for these reasons.
Local data recovery is often more complex because an alternate disaster recovery location is often needed above and beyond the built-in tools in a storage system. This can mean buying twice as much hardware and supporting a separate DR site. Understanding the needs of the business is crucial, as long-term backup and failover options may be needed in addition to local snapshots and basic operational recovery at a single local site.
10. Support
Consider the level and form of support you'll need before you choose a cloud provider. The most important options vary among providers. They include 24/7 support and response time, tools to monitor and report on your environment, consultative agreements, dedicated account managers and engineers, and proactive reviews with subject-matter experts.
Local SAN architecture has the same considerations, but you must ensure there are enough staff members with the appropriate training to support the infrastructure, as well as the appropriate support contracts in place to get help from vendors when needed.
11. Vendor lock-in
Lock-in is a consideration for both cloud and local deployments. When assessing a vendor, review your migration goals and current infrastructure costs and resources, and look at each vendor carefully. You should have an exit strategy because data lock-in is the most difficult risk to mitigate, especially in a cloud deployment where it's difficult to switch vendors. Set up your data for maximum portability.
12. Manageability
Manageability can underscore the challenges of cloud storage vs. SAN. Cloud services can reduce infrastructure manageability. For instance, direct access to server consoles and direct control over what's running on shared infrastructure is limited. All service providers support different orchestration tools and integrate with different services. An organization must consider what level of control it requires over its environment.
In a local SAN deployment, vendors generally offer management tools. There are also numerous vendor-neutral options. It's critical to do demos and find what suits your needs.
13. Reliability
Cloud computing enables IT staff to rapidly develop complex systems and deploy them across the globe, but this approach can create reliability risks. Cloud providers are continuously developing new ideas and including them in their services. The rapid rate of change can cause service failures. Consider globally redundant cloud offerings for maximum recoverability and reliability.
Local deployments require more planning and management to ensure business continuity, generally with multiple sites and redundant hardware. Options to consider include replication to third-party cloud vendors and the creation of a local business continuity plan with redundant hardware at a privately owned alternate location.