Has Starwind lic. it's tech ???

[fbifido]

Hi,

not sure if I can ask this question openly
but I just say a video from a company that VMware purchase
virsto[dot]com

They call their tech Storage Hypervisor
it install on the nodes and on the storage server.
it uses vLogging with normal hard drive, with inline-dedup. and other similar tech I have read on starwindsoftware.com, I did not however see any tech on logging file system.

Question:
When I first saw that they install stuff on the nodes(hypervisors), I ask myself "Is it possible to reduce network traffic between the hypervisor and the Storage using compression or inline-dedup(if possible) of the network traffic itself ?????"

BTW:
1) can't wait to test v8, it will work with server 2012 R2! Yes ?
2) Can't seem to find mush info or video demo of upcoming v8 on your website
3) I would like to do clustering, step-by-step, as in start of with 1-hypervisor and 1-storage, then add another hypervisor, then another, then another. Also add another storage.
3a) is it possible to use free space from both storage server as one iSCSI target, like say scale-out-iSCSI, LOL ?
or even raid across storage server?

[Max (staff)]

Hi Fbifido,
I can proudly say that all StarWind solutions are developed in-house.
As for your question:
The only way to minimize the network traffic between the hypervisor and the SAN is to use DAS.
Alternative way is to carry out deduplication and compression between VM layer and networking layer.
However, this doesn't guarantee data integrity in case of power outages. To maintain it the data still needs to be sent somewhere.

1. StarWind v8 works with Windows Server 2012
2. There is no problem in finding information which doesn't exist. We provide limited information about the upcoming version, stay tuned for the updates.
3. That is something you can do with StarWind.
3a. It is possible to do all of the scenarios you've mentioned. Keep in mind that you either need to attach this storage to the cluster or put NFS on top since block level storage only likes clustered FS or NFS on top of it. An initiator server which isn't aware of others actions with the iSCSI storage device can corrupt the data with a single write operation.