BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
Evaluate
Weigh the pros and cons of technologies, products and projects you are considering.
Hyper-converged system checklist evaluates required features
With multiple hyper-converged systems to choose from, this checklist can help you determine which system is right for you.
With all the hype over hyper-convergence, and the varying definitions from vendors, it's important to know exactly what you're looking for in a hyper-converged storage system. Will you need to purchase new hardware? Does your IT staff need training from the vendor? From potential hardware issues to manageability and support, a hyper-converged system should meet all your storage requirements.
Storage expert Jon Toigo has created a handy checklist of questions you should ask when evaluating hyper-converged system options.
1) Hardware dependencies
- What hardware will I need to make the offering work?
- Can the components be upgraded independently?
- Which vendor(s) sell the nodes/equipment/components for this product?
2) Multi-nodal architecture
- What is the recommended number of nodes for high availability and performance?
- What is the total cost of a node (hardware + software)?
- Are all nodes in a cluster required to have identical configurations?
3) Manageability
- Which storage services will be used with the offering?
- How does the product scale storage capacity independent of compute?
- Does the product work with other hypervisors as well as non-virtualized environments?
4) Hardware utilization efficiency
- Does the product make the best use of DRAM and flash?
- Does the offering require you to retire your existing storage investments before you are ready (i.e., rip and replace)?
5) High-availability clustering
- Is nodal clustering functionality integral or is it provided via the operating system, hypervisor or third-party clustering software?
- How is data mirroring configured and administered?
- Are memory caches and buffers mirrored between nodes?
6) Mirror test and validation
- How are mirrors tested?
- Can mirrors be verified before failing over between nodes?
- Is nodal failover an automatic process or is manual intervention required?
7) Metro clustering
- Does the offering support stretch or metro clustering across a WAN or metropolitan-area network?
- Is synchronous and asynchronous mirroring capability included in the solution?
- Do local and remote nodes need to have identical hardware?
8) DRAM support
- Is dynamic RAM supported for buffering writes?
- How is a DRAM buffer protected from data loss?
9) General fitness to purpose
- Is the hyper-converged system suited to the environment in which it will be deployed in terms of power consumption, cooling requirements, noise requirements, physical space available?
- If staff training is required, is it available from the vendor? At what price?
- Will the offering support your workloads/hypervisors?
This was last published in May 2015
Join the conversation
1 comment