As our clients migrate applications and services to Infrastructure as a Service (IaaS) platforms, they are facing some fundamental questions related to ensuring that their applications and data are both available and protected. The West Monroe Cloud Operations Reliability Engineering (CORE) practice has assembled a number of methodologies and best practices regarding the selection, implementation, and operation of IaaS-based infrastructure solutions.
Due to the "cloudy" nature of IaaS solutions, the true capabilities of the built-in high availability and data protection features can be difficult for inexperienced system administrators to design and implement effectively.
The following include a sample of West Monroe best practices to select, design, and operate hosted infrastructure solutions:
IaaS provider selection tips:
- Select an IaaS provider with a solid track record of transparency and proactive communication of technical issues and their resolutions.
- Obtain and evaluate the technical details on the major platforms and infrastructure systems deployed as part of the IaaS service, and their provisions for high availability, monitoring, and disaster recovery.
- Ascertain whether or not your IaaS provider has the correct systems management tools, procedures, and expertise.
- Evaluate the IaaS provider's SLAs and technical details for system recovery, in the event a localized hardware or software failure interrupts the services they are providing.
IaaS-based system design tips:
- Assume that your IaaS provider's infrastructure will not be reliable, and design the full IaaS solution with high availability and disaster recovery in mind.
- Depending on the projected impact of an IaaS outage to your business, consider leveraging multiple IaaS providers and replicating the services and data between them.
- Consider a hybrid IaaS design - allowing you to replicate services and data to your company's internal systems, while maintaining copies of critical data separate from that of your IaaS provider(s).
IaaS-based system operation tips:
- System administrators need to stay engaged with the IaaS provider’s evolving service offerings and known issues, adjusting services/configurations as necessary.
- Systems and applications deployed on an IaaS platform should be monitored for availability and performance through a system/tool separate from that of the IaaS-provider's.
Depending on the size of the environment to be hosted and its importance to your business, not all of the above recommendations may apply. Understand your business and operational requirements prior to designing and implementing an IaaS solution, and incorporate the proper CORE technologies and processes into the solution.
What have your experiences been with designing a reliable IaaS solution? Are you finding that IaaS vendors are responsive to your functional and operational requirements for reliability and data safety?