Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

IIIS 2013: Technical Workshops

T.C. Seow | June 28, 2013
Reports on the Technical Workshops held at this year's Implementing Information Infrastructure Summit in Singapore.

As to what cloud storage is used for, there are several reasons. Elastic demand for Web-based media (such as video, ebooks, audio), backup to the cloud (restore, recovery, "seeding" the backup with hard drive), synchronisation of files to the cloud and multiple devices (Internet "drive" as secondary storage), archive/preservation in the cloud (including compliance, retention and e-discovery), are some of the main reasons warranting cloud storage.

He also talked about the data storage interface evolution, which referred to the huge amount of data generated by an ever diversifying set of devices such as mobile phones and tablets. With trends like BYOD and consumerisation of IT, corporate data are being stored in public cloud in unmanaged and unprotected states. This might be prone to be compromised.

All these changes to storage requirements are leading to the increasing importance of cloud storage. SNIA, together with a number of standards bodies, are working on the so-called cloud data management interface (CDMI) to present a vendor-agnostic way of accessing and utilising cloud storage. The key advantages, Chua said, are that no one particular vendor would have absolute control over the interface, and cloud vendors would not feel threatened that someone else might be controlling their interface.

"CDMI is under change control of a standards body, accommodates requirements from multiple vendors and can be extended for proprietary functions," said Chua.

The technical session spent some time on the adoption rate of CDMI, and ended with some detailed treatment of how cloud storage could be deployed, and what technical issues should be addressed before deploying.

Realities of solid state storage
Heoh Chin-Fah, who is also an education Instructor with SNIA South Asia, took the stage to talk about how best to take advantage of flash memory in enterprise storage environments, in particular in recognising the relative advantages of flash tiering, caching and all-flash approaches, with reference to performance, cost, reliability and predictability.

After giving a simplified description of flash memory, Heoh went on to explain its inner workings, and why flash "wear" is a real concern in how it should be used within a storage infrastructure. "In the end, capacity, cost improvement brought about by flash, and the attractiveness of its form factor must be weighed against true concerns such as flash performance, reliability and error-handling," said Heoh.

Heoh Chin-Fah
Photo: Heoh Chin-Fah

Heoh went on to talk about the technical differences between flash and disk drive memory, fleshing out the key properties of each drive technology. While flash wins in access, Heoh warned of flash "read wear", an inherent shortcoming of solid-state architecture that does not suit high-speed random read/write input-output processes. Disks on the other hand spend a lot of time seeking and rotating (the disk), compared to data delivery.


Previous Page  1  2  3  4  5  Next Page 

Sign up for Computerworld eNewsletters.