Yesterday I drove up to San Francisco in the middle of a record-setting heatwave to attend a dockside party thrown by Zuora, a billing-automation SaaS startup.  They were celebrating signing their 100th customer, as well as signing Sun, who hopes to compete with ENKI in the Cloud Computing arena sometime soon, using their recent aquisition of QLayer.  I was there to look at their software, but also to talk to the people. 

Schmoozing about in the heat, I met a few bright-eyed startup entrepreneurs as well as a preponderance of Zuora partners.  I got a demo of Zuora, which is a very snazzy billing package that allows for all manner of subscription and on-demand billing to be automated.   ENKI needs something like this, since our NetSuite ERP system doesn't have the faintest idea of the billing challenges we face.  In addition, we are considering offering billing services to our customers who themselves run SaaS applications.   We've held off on Zuora ourselves in part due to the price, but also because we wanted to be sure we understood the problem before looking for a solution.  I understand it now... all too well!

What impressed me - and explained the number of partner companies at the event - is that Zuora has really tried to focus on one thing alone: billing.  To do this, they have to own their customers' item list, which means that traditional  business applications like ecommerce and accounting have to be integrated with Zuora.  After implementing a SalesForce.com integration, they switched to relying on Boomi for their integrations to get more traction.  As a result, they have come a long way in a short time, by focusing brutally on what their value is raather than integrating with everyone else.

This, I think, is the lesson for business in these times: focus, focus, focus.   For us at ENKI, it means that we'd be best off leaving billing to someone else even though I shudder at the price.  For you, our customer, it means that if you can avoid coding by integrating with other services, you should do so.  That way your product will make it to market faster, and your time to value will decrease, which will bring you investment or revenue as you have planned for.

These were the amazing words I heard today from Larry Halff, founder of Ma.golia.com.  I was submitting a blog article today and decided to check out the little link sharing icons our website puts below each article to submit it to link sharing sites.  I tried to go to ma.gnolia.com, one of my favorites.  To my surprise, it was gone, though their cute flower logo was still visible.  Instead there was a web site with a painful story of file loss and implosion of their site, with a very interesting video from the founder of ma.gnolia.com listing his mistakes and learnings.

Citizen Garden Episode 11: Whither Ma.gnolia? from Larry Halff on Vimeo.

In short, their MySQL database got its files corrupted.  Their half-terabyte database was backed up, but the corruption was backed up as well.   Eventually, the data became unreadable, and ma.gnolia.com came to and end.

The founder, Larry Halff, showed tremendous and admirable humility in listing his mistakes, but he also told a cautionary tale that we here at ENKI both lived ourselves at startups as well as observed with some of our customers prior to their joining us:

1. MySQL is a dangerously inadequate database for production work.  Sure, YouTube uses it, and so do hundreds of thousands of LAMP sites.  But it's very susceptible to corruption, and its administration facilities are inadequate to prevent it.   In the hands of inexperienced or amateur DBAs, it is ripe for disaster, yet it is presented as an easy entry-level database.   We've had customers suffer long downtimes due to corruption from crashes in MySQL.  We also have customers with nearly insoluble performance problems.  And, even though our cloud computing environment restarts software on failed servers automatically, MySQL does not always successfully survive a restart, or if it does, it requires extensive database table repair (especially with myISAM tables).  If uptime is important, we feel you should look elsewhere, though it can be made to work with sufficient expertise.
2. Startups shouldn't do their own IT.  I bring this up often because of my own experiences as a software person trying to do my own IT, or watching the stream of missteps in the early years of startups I've worked at including NetSuite.  In the video, Larry says, " The real lesson Learned is if you’re a startup, don’t do your own IT at all.”   I'm hoping it's a lot more convincing when it doesn't come from my mouth, now that I've started a company to address this need.
3. Infrastructure is what makes the difference between a web site and a web business.  Larry points out that "in the process of developing Magnolia, infrastructure always took a back seat."  Then he and the co-host go on to joke about how Cloud Computing (and Amazon) would have solved their problems.  However, the actual failure they had would have happened in Amazon AWS just as well as their homebrew hardware, because the root cause was a lack of IT experience at ma.gnolia.com, not having physical hardware versus the Cloud's virtual hardware.  What Larry needed was IT expertise expressed as IT practice and procedures, yet he didn't have any (and shouldn't have had to learn it himself since he was the creativity guy!)

Since announcing automatic allocation and scaling of cloud resources in January, ENKI has been selling and providing customers with the ability to have their applications monitored by our monitoring system, and automatically have additional resources added and removed from their virtual private data centers to conform to varying loads and to save them money.  I wanted to share an update on this with you and show you some pictures!

Customers who can benefit from this must have an application in which additional instances of virtual servers can be added and utilized by the application software: for example, additional web, application, or other data processing servers that are handed work by a load balancer or job scheduler.  ENKI offers reserve capacity on an as-needed basis, or accepts reservations for reserve capacity so that it is guaranteed available when needed.

The mechanism we use to effect the auto-allocation is a controller node inside the customer's virtual private data center that can access monitoring information either for virtual servers or the load balancers that send requests on to them, and is authorized to request resource provisioning and disposal from AppLogic.  The policies for allocation and disposal can be set uniquely for each customer to optimize their load response profile to their business needs.  Because AppLogic has a persistent model for virtual instances, there is no need to allocate a new instance, but rather only turn it on, which speeds deployment to less than 30 seconds, versus other cloud vendors' potential 10 minute delays.  This makes the application more responsive, though it doesn't eliminate the need to size the "first" instance correctly for base loads.

The good stuff - a case study

We have one customer who serves ads for people and companies wishing to advertise on electronic greeting card sites.  The greeting card companies auction off blocks of ads, and then give the ad services company (ENKI's customer) the responsibility of serving the ads when their bid is accepted.  This creates a widely varying load on the ad services company, which is an ideal opportunity to take advantage of auto-allocation.    In this case, the customer has three virtual machines that serve ads, called A, B, and C.   A runs all the time, serving the base needs of our customer's business.  When bids are accepted, the controller watches the load on A go up, and turns on B when a predefined threshold is reached.  Similarly, if B becomes too busy, C is turned on.

Now for the pictures.  You can see the load (as total network traffic on the load balancer) in the first graph, with daily peaks corresponding to people's electronic greeting card usage habits.

Load:

Below are usage graphs for each of the three servers as they serve the load shown above.   The first graph shows the load on the application server which is always running. As this server, A, peaks, our Cloud Auto-Scaling system activates additional servers. 

Server A:

In the graphs below, gray areas indicate that the server is off, so our NimSoft monitoring system cannot measure the load.  It wakes up when the server does, and shows its load while it is running, until it is turned off automatically again.  You can see how some load peaks cause B to be turned on, and a smaller number are large enough to activate C.

Server B:

 Server C:

What we see from these graphs is that there is some cost-saving optimization left to be applied: the primary server, A, is still idle a lot of the time.  Decreasing the resource allocation to A and perhaps increasing B or even making C twice as big as B could potentially save our ad-serving friends some money.   This really illustrates the power of in-depth monitoring when added to auto-scaling technology in the cloud: it produces analytic performance analyses that can allow businesses better control over cost and performance tradeoffs. As it currently stands, this customer is saving about 33% over simply allocating for peak usage.  With some tuning, they could save up to 50% or more.

I wanted to share my thoughts on how ENKI can assure our startup customers of success with the operation of their live sites, and hence their enterprises.  It's what we were originally founded to do, and I'm seeing from our current customer base that the world of startups still desperately needs the experience, process, and tools that we bring to the table to assure that software deploys reliably to your live site and runs successfully.  Of course, these requirements apply to enterprise deployments as well!

Going back to basics, lets talk a bit about the problem that we're trying to solve, and why it's interesting.  From the 50,000 foot level it looks like your startup being unable to meet its go-live dates, having problems with post-release quality, and uptime problems.  I watched this happen in the early days of NetSuite and other companies I worked at, as well as some of ENKI's customers, and most of them took a long time to iron it out. This situation, if it occurs, really revolves around skill sets that a typical software startup has.  There are lots of smart programmers and problem-solvers which can bang out an interest-grabbing new product in record time.   However, what's often missing are the skills and experience necessary to do the boring work of repeatably releasing quality software on-time.  Companies that are successful with this effort have shown that the necessary skill sets are:

• Project Management
• Structured Design and Coding (including revision control)
• Structured Software Test and Inspections
• Production IT (focused on fast and reliable services)
• Business Process Management 

This is where ENKI comes in: our founders and staff have already done this in and for many companies.  We can assist with the production IT and BPM, as well as provide facilities to make project management, structured design, and structured test easier for our startup customers.  While we've always offered these services, I felt it was time to package them up and offer them as an agreed-upon way to describe how we work with our customers.

I'm calling this bundle the Startup Success Program™ since it's a package of services, processes, and technology, delivered in a partner relationship between ENKI and you, that accomplishes this goal.  The overall framework for the Startup Success Program™ (SSP) is a structured release process that looks like this: 

To support this process, the kit will contain:

These services are available a la carte from ENKI, but I'm hoping that delivering them as a single contracted offering will allow us to build a relationship with our startup and strategic customers that lets us assure them of a given service level as well as adding true value for their end-customers.  It also avoids the uncomfortable "well if you had ordered this services then we could have avoided that problem" discussion which places the "blame" for operations problems onto people who shouldn't have to be responsible for the details of avoiding them in the first place.

Finally, for select startups and strategic partners, we shall offer the the SSP with a 50% discount (60% in 2012!) from base price on computing charges (CPU/Memory) for the first year, as well as discounts on storage and operations services labor, which helps to overcome many startups' challenge with pre-money financing and first-customer investment.   To receive this discount, contact us and share your business plan with us (under NDA if you wish) so that we can evaluate the strategic benefits of a partnership to both parties.

Postscript: check out the discussion about the fall of ma.gnolia.com due to mistakes made in IT in my blog article, don't do your own IT at all.

Compare the SSP price (red curve) to ENKI's regular pricing: