This means PolyServe file server nodes are mapped to a shared volume on a SAN, so if the SAN doesn’t scale-out like the DFS, then the SAN becomes a bottleneck.
PolyServe on EMC, for example, does well but hits an I/O bottleneck at the EMC controllers. Having a scale-out file system on a scale-out block storage system is the “holy grail.” Each file server node is mapped to a shared and scalable volume on a SAN.

Ryan, LeftHand is still taking orders and will continue to support the Dell and IBM based storage server platforms. If I were to guess, HP would eventually like to offer an HP equivalent of the Dell and IBM offerings that can be added seamlessly to the Dell and IBM based storage clusters. LeftHand’s customer support organizations will stay in tact and continue to improve. I believe HP sees this as one of LeftHand’s strengths, so why change what’s already working.

BTW, I’m out there beating up on EVA all day long and haven’t slowed down one bit; let the best product win!

My fear is that the HP acquisition will mean the elimination of support for SAN/IQ on Dell and IBM platforms, and artificial limits on high-end features and scalability to protect HP’s EVA lineup.

Also, LeftHand’s premier support is absolutely fantastic, better than any other vendor we work with. My experience with HP’s x86 server support organization has been mixed at best.

HP has a history of destroying their acquisitions – I hope LeftHand is an exception.

XIV (pretty limited flexibility) and LeftHand Networks, very slow growth over 9 years in business, are both block based storage products.

The rest are clustered file system solutions. File based access vs. block access meets two different needs. From my experience in the CFS market, the driver for CFS is both single file access as well as aggregate access to the same file space.

LeftHand Networks will never be a Clustered File System, Parallel File System, or a Networked file access solution.

They will be a scalable iSCSI software group within HP…and I hope for the people who are working there, that they are treated as well as Dell treated the Equallogic folks. LHN is in my backyard, and while I’ve always been aggressive in the marketplace against them, they do employ many of my friends, and i want nothing but the best for them.

As for the LHN deal? 4x revenue is a great offer for any company, which is about where I place HP’s offer on LHN. VCs didn’t kill it, but they didn’t lose the investment either. As far as profitability, that is all a game of shuffling money around, companies that are looking to IPO or be bought will invest less internally in order to look more profitable, or show major investments in engineering to make up for less profitability.

The real question I have for HP, will they be able to pull LHN into the enterprise storage space? or will they just let it simmer in SMB so that the EVA revenue stream isn’t impacted. I think that is the biggest difference between Dell’s buying Equallogic and HP’s purchase, Dell’s low margin EMC business (in the enterprise) is the 800lb. gorilla that no one wants to think about other then EMC (since that revenue isn’t low margin for EMC). Dell only gains from Equallogic’s growth, HP might suffer (never looked at the margins on MSA and EVA could be interesting). As for LHN’s margins, no way to know, since they have never formally filed with the SEC, it is all speculation.