I talk to me about HPE

(This blog post was also published on LinkedIn)

Let’s get the disclaimers out of the way. HPE paid for my travel, lodging and food to attend HPE Discover 2016 earlier this month. That’s it (besides me having been an employee there in the past), they aren’t paying me for my opinions, writings, rants or videos.

I see a lot of good things going on at HPE. They say they deliver solutions to help customers IT organizations’ become more agile and they appear to be eating their dog food. For example, after they tried to compete in the public cloud business, they decided to focus instead on producing their excellent Helion private cloud management software.  They also adopted flash technology in their flagship 3PAR storage product line in a very effective, straightforward way, as opposed to creating confusing and diverging product lines the way EMC and Netapp did.  It appears they are even starting to figure out how to leverage the technology and team from their disastrous Autonomy acquisition to develop enterprise software for Big Data and IoT. In short, they seem to have figured out the markets that are important to them, the products they can sell today and the investments they need to make to compete in the future.

The pendulum of focus appears to have swung to the technology and product side of the business and away from marketing side. This was necessary, but HPE also needs to figure out how to communicate effectively about their technology and products, which is not easy for a company suffering from branding/naming confusion. Names are a tough challenge for many large IT vendors and HPE often struggles by inflicting good technology with unfortunate names – including flagship tech like “Composable Infrastructure”. The word composable does not mean anything to anybody and dictionary definitions shed no light whatsoever on what HPE is trying to communicate. This does not lead to a sense of mystery and capability as much as a sense of baloney. Ambiguity will not help HPE make it to where they want to go and they need all the clarity they can get.

HPE is not back to where they used to be because there really isn’t a “back there” anymore – the world has moved on and HPE is much better situated to pursue enterprise technology opportunities than HP ever was.

Qumulo Emerges and Nudges the Awareness of Data Awareness


swimmer surface tension

Last week was good for the storage universe, as Qumulo announced their data-aware NAS systems called Qumulo Core.  Qumulo stole a page from DataGravity’s playbook by positioning Qumulo Core as data-aware storage, but the two products don’t seem competitive – Data Gravity sells its systems mostly to small and medium sized customers, while Qumulo appears to be targeting very large customers.  Here are a few links to what others are saying about Qumulo Core:

StorageBod (is mostly interested in lower cost competition to Isilon)

Stephen Foskett (likes what he sees in their QSFS file system)

Scott Sinclair from ESG  (wonders how companies will adopt this new technology)

Will it make it and what will it look like if it does?

turtle hatchingThere seems to be plenty of speculation about what Qumulo and data-aware storage will become. Some seem to think that data-aware is just the latest buzz word from the storage industry while others think it could have profound impact on the ways data and storage are managed. Unfortunately for Qumulo, there seems to be no clear recognition of the big problem that Qumulo Core supposedly solves. It does interesting things like tracking IO activity levels for individual files, but the question is why should this matter?

The answer is that Qumulo will matter if customers can save a lot of money by installing Qumulo’s storage – regardless of the value their data-awareness technology. The stickiness of their products however, will depend on the value of data-awareness. If people start managing their storage differently and more efficiently with Qumulo, it will force changes to the rest of the industry.

It’s easy to say you have cheaper hardware than the competition, but changing storage platforms typically involves migrating data. If customer data has to be migrated from an existing platform to the new one, there is a lot of planning and administrative work that needs to happen, which raises the cost of the new solution considerably. Greenfield opportunities are the best for Qumulo because customers can pocket the storage savings without regard for data migration costs. But greenfield opportunities are a tiny subset of the enterprise storage market and Qumulo will not succeed if it relegated to that niche. saving turtles

Who will help them?

Most startups need help selling their products. This is particularly true for lower-cost storage products where the cost of direct sales can be prohibitive – mandating a channel sales model. I expect Qumulo will attempt to recruit Netapp resellers who need products to compete with Isilon. In that scenario, Qumulo could become a pseudo-friend to Netapp by becoming the enemy of their enemy (Isilon/EMC).  However, that friendship would not be all that close, considering Qumulo can’t afford to turn down business from Netapp accounts. Not only that, it’s not clear how many Netapp resellers also sell Isilon already and don’t need a competitive product.  The Go-To-Market strategy Qumulo has will be very important to them as they find themselves competing for deals through top end storage resellers who are also working with EMC and Netapp.

Storage has never been about building a better mousetrap, it’s about competing in every facet of the business. Qumulo is going into the teeth of the enterprise storage beast just as many others have gone before, some successfully and many others, less so.

Starting up at Quaddra Software

mountain-bike-jumpThere’s something about working for a startup that gets my motor running. There’s more at stake (more thrills) and more freedom to do things that don’t fly in large, organizations. So here I am, back at a startup, Quaddra Software, after a year and half with Microsoft.  Microsoft was a good place to be an employee, but I’m a “journey-guy” and am more interested in overcoming new obstacles than repeating old ones.

Quaddra is developing highly-scalable, high-performance file analytics software that leverages open source software with a pluggable architecture for adding new functionalality. I really like the potential of this technology because its value comes from creating intelligence from opaque data and creative customers always find interesting new ways to use intelligence.

Timing mattersthings have changed

Beyond the concerted efforts of talented people working together, two of the most important elements to a startup’s success are luck and timing. Startups always need as much luck as possible and should do all it can to create its own. Timing is one of those cruel things that a startup can’t do much about. If it’s too early, the company has to create a market by itself without an ecosystem to support it and the odds are very good that the company will run out of money and go out of business. On the other hand, if the startup is too late, other companies’ products get the best opportunities and the startup suffers from slow growth and lower margins – a slower path to going out of business or becoming one of those zombie companies that is not really alive, but not yet dead either.

The best scenario for a startup is to have technology siblings that compete and grow together,  increasing awareness of the solution set and expanding the market much faster than they can on their own. It’s that old strange math where 1+1=3.  10062010_catalina1 (2)This go round, it appears Quaddra will be sharing a room with a company populated by friends from my EqualLogic days Data Gravity. From reading their blog and with their public announcement today in the Wall Street Journal, it appears we are on very similar vectors. I have a lot of admiration for Paula Long, John Joseph and the rest and I hope they are wildly successful and that Quaddra competes with them in the file analytics business for a long time.

Applications for Quaddra Software’s

IT teams that know a lot more about their company’s unstructured data can control it, manage it and generate reports about it.

Control – Quaddra’s software scans and searches unstructured data in-place where it resides and builds independent indices where system administrators search, identify and tag files for various actions. For instance, certain media file types that are not be part of the normal corporate work streams could be identified as non-work files and removed from corporate storage. Likewise files with sensitive data could be found in locations where they might pose a risk for loss or leakage.

Manage– Quaddra’s software can copy, migrate and archive data from it’s current location to to virtually any secondary store, including cloud object storage. Files can be identified using many different criteria and acted on according to management policies that safely store historical files and make free space available on primary storage.

Report – The results of searches can be used to generate reports to analyze and share with managers and co-workers to make informed decisions. For instance files can be analyzed by their access data and correlated with the amount of capacity they consume to determine candidates for migrating to archival storage.

age analysisIf you want to talk to us about developing intelligence for your opaque, unstructured data, please send an email to: info@quaddra-sw.com