Virtual Flash (vFlash) Tech Preview

"Those of you who have read the technical preview article on Distributed Storage will have read about SSD being used as a cache (for both read and write I/O)"

First off , the scale we are tying to relate to is in the server environment for the redundant backup of stored data. That is how ZeroSnap is iitially presented.

"SSD being used as a cache"

Using the SSD in that manner, if considering the memory issues only and not its on board I/O controller principles, is more used as a non-volatile on-hand use of data.

I'd compare that to a less scale of L1 or L2 volatile RAM for on hand data cache more familiar in a PC environment.

Both instances are for data that is accessed frequently.

Utilizing L1 and L2 cache (RAM, random access memory) such as within a PC environment, will not require any specialized I/O principles to place data within them read /write.... with that, the memory is volatile.

Utilizing the memory of the SSD as a cache will require a controller of some sort to write / read the memory. Cache is normally associated with data needing fast random access.

What e.Digital defines as cache is not the flash memory or cached data on it, but the I/O controller principles interfacing the flash memory. These I/O controller principles are designed for various uses...and it allows for the emulation of Random Access....without the need of substantial RAM resource when combined with the MOS.

It is noted that the SSD as a cache ....."The vFlash infrastructure does not specify which Virtual Disk data is to be cached in vFlash. That decision is left up to the vFlash Cache Module."

Once, whatever metadata is compressed(see my prior posts on ZeroSnap) are physically written into the flash.....how do they get a simplified random access normally associated with the meaning of cache data utilizing flash memory?

No matter how overly involved the server management environment is....RAM is ram....FLASH is flash....at any scale.

The principles of ZeroSnap and vFalsh are for one purpose....shrink the overhead of legacy virtual environments.

Shrinking the overhead and managing it in front of: "With VM-aware caching, chunks of flash can be allocated on a per VM basis, and these will show up as SCSI disks in the Guest OS"...... is the what I question.

anyway

doni