[LSF/MM/BPF TOPIC] NVMe HDD planning and directions

[LSF/MM/BPF TOPIC] NVMe HDD planning and directions

[tim walker]

Hi all-
I'd like to propose we talk about the PCIe-NVMe HDD this year. The 
topic comes up informally from time to time, but now that we have
products with actual delivery schedules I think it would be valuable
to discuss the impact to our systems. 

Background:
NVMe specification has hardened over the decade and now NVMe 
devices are well integrated into our customers’ systems. I think
moving HDDs to the PCIe/NVMe interface is the next logical step. 
Consolidating on a single API for rotational and static storage 
technologies. PCIe-NVMe offers SATA-level costs while supporting 
features and performance suitable for high-cap HDDs, and optimal 
interoperability for storage automation, tiering, and common 
management tools. 
We will share some early integration results, and review the
OCP design guidelines, but really don't want to turn this into
a status report or marketing blitz. HDDs, PCIe, and NVMe are 
all mature technologies, but the combination is new. We are 
looking to the experts to help us anticipate the challenges 
and benefits.

Discussion Proposal:

-What Linux storage stack pitfalls do we need to be aware of as 
we field these devices with drastically different performance
characteristics than traditional NAND? For example, what 
schedular or device driver level changes will be needed to 
integrate NVMe HDDs?

-Are there NVMe feature trade-offs that make sense for HDDs 
that won’t break the HDD-SSD interoperability goals?

-How would multi-actuator HDDs be presented under NVMe? 
Namespace-per-actuator?

-NVMe HDD power management proposals. This might 
turn out to be the hardest part.

-NVMe HDD System architecture (maybe, if there is interest?)

Thanks, and best regards,
Tim Walker