Ah, the buffet. A dining experience where anything you could want is served to you from a magical kitchen that makes an endless feast of delights.
Today, the Datanauts bring you just such a buffet. It’s a smorgasbord of storage, and we assault and pepper our guest with all sorts of questions.
Howard Marks, chief scientist at consultancy DeepStorage.net and cohost of the Greybeards on Storage podcast, answers all our inquiries about the latest in storage technology, including flash, cloud, object storage, and virtual volumes.
Viptela’s hybrid SD-WAN technology unifies your MPLS, Leased Line and Internet-based WAN infrastructure. It’s packaged with centralized administration, real-time visibility, and policy control for precise and measurable application performance.
We’re recording a live Packet Pushers Podcast with Viptela at the ONUG conference this November. Join us November 4th to meet the Packet Pushers and hear from Viptela customers about real-world SD-WAN deployments.
Part 1 – Array Architectures And Flash
- Do all storage arrays use flash in some part of the architecture?
- What are the main reasons one would choose all flash versus a hybrid of flash and disk?
- What are the differences among the flash devices being used inside of storage arrays – traditional SATA-based SSDs, PCIe flash, NVMe, DIMM based, and so on?
- I’m using a flash array in production, and need to replicate data to a secondary site to meet my RPOs. How do I justify another flash array for DR, or should I go a different route?
Part 2 – Cloud And Object Storage
- What types of storage are available in the cloud?
- What is object storage?
- What makes cloud storage attractive – either financial, workload specific, application specific, or geographically advantageous?
- Why is object storage so common for cloud storage providers?
- What is the difference between file, block, and object storage?
- What are the differences between eventually consistent storage and immediately consistent storage – I’m thinking of traditional high-performance block storage, versus geographically dispersed object storage.
- How do you avoid a Hotel California scenario – where I have so much data in a public cloud that it’s nearly impossible to get it back without paying two arms and three legs? Or do I care?
Part 3 – VMware Virtual Volumes (VVols)
- VVols seem to be something unique vs. traditional block and file consumption. What are Virtual Volumes?
- What sort of workloads would benefit from VVols?
- Is VVols something that helps block IO blenders, file storage, or both in some way?
- Why has it taken so long to get VVols released, and how are storage vendors reacting to the release?
- If I already have a storage array, and it’s not going anywhere for a while, is there any way I could take advantage of VVols – or am I sunk until it’s time for a renewal?
- How does VVols compare to VMware VSAN storage?