Oracle to Amazon: It’s on

I worked for Oracle for a little while after getting acquired.<p>For a developer, there are so many obstacles you need to overcome to ship anything on time there or of quality. Everything is bad, from the developer hardware&#x2F;tools to the red tape.<p>The bug tracking system looks like an HTML 1.0 web page. The &quot;database version&quot; is a required field, even though the product you work on doesn&#x27;t even use a database. I seem to remember having problems linking to issues, because the the URL was just a hashed session so you couldn&#x27;t even bookmark things.<p>You want to use that open source library that everyone else is using and does exactly what you need? All you need to do is fill out a bunch of on-line forms where you explain the material impact of using this library and how much money your BU makes and how this affects the bottom line. Once you get through all that, your request needs to travel all the way up the food chain to the office of the CEO which can take months of harassing administrative assistants of very highly payed people for that to actually happen. This doesn&#x27;t even take into account the legal approval steps that need to happen which also takes a while.<p>While this is going on, you have to work with your 3rd party library and hope that it gets approved or risk months of wasted work. In the end, you notice that other people don&#x27;t even try and use 3rd party libraries because it&#x27;s so frustrating. In the end they just try and re-invent the wheel, and because a certain percentage of your team has to be in a &quot;low cost center&quot;, the library ends up sucking.<p>Recap: Amazon has nothing to worry about, because nothing innovative or powerful is going to come out of Oracle.

I did not realize that Oracle was in this much trouble. In my experience companies that are circling the drain put out press releases that say &quot;Hey &lt;leader of some technology that is hip&gt; we&#x27;re going to take away all your business with our version of your same technology that we just figured out how to build, sort of.&quot; Stepping into this mess is not something the Oracle organization has the DNA for.<p>I grant you my experience with them is dated (they were a huge part of the NetApp push into data base markets) but they were all about the &#x27;added value option&#x27; and high touch management &#x2F; maintenance. Not something cloud compute is about.<p>To my way of thinking it seems like step one here, would seem be build a PUE &lt; 1.05 data center with 10MW of provisioned power. Have they done that?

&gt; <i>the real dough lies in Software as a Service (SaaS) and what Ellison called a “highly differentiated” Platform as a Service (PaaS) that will provide it Oracle-like margins on enterprise applications.</i><p>I call bullshit. No one is making &quot;Oracle-like margins on enterprise applications&quot; in the cloud...yet...not anytime soon in this competitive market.<p>This is the opportunity in IaaS&#x2F;PaaS as I see it: every large company in the world typically outsources their operations systems infrastructure. Here&#x27;s an example - a company based in Europe that does €7b in revenue. Let&#x27;s call them Big, Co. They have a myriad of technology partners - Oracle, SAP, HP, IBM...you name it, they have it. Anyway, one of their partners, Atos, provides infrastructure services for them based in Austria. So, let&#x27;s say the CIO decides that the HR department is going to use SAP HCM to support HR. Some manager at some level calls up Atos and says &quot;Alright we need 6 HP ProLiant servers, Oracle 10g installed on 3 of them, and SAP HCM installed&quot;. This process takes months. To everyone in the web world who&#x27;s unfamiliar with enterprise - this is the equivalent of installing a LAMP stack on EC2. So you can see the disparity between the two (months vs minutes). The problem is - none of the enterprise technologies (Oracle 10g or HP ProLiant blades for example) are sold as commodities...they&#x27;re &quot;professional&quot; and &quot;enterprise grade&quot;. So here&#x27;s the opportunity - do what Amazon did for the web in the enterprise. So, let&#x27;s say you want SAP HCM, well that&#x27;s simple, have support spin you up a server with a pre-installed set of technologies (like a LAMP stack!) and then let some implementation managers go and install it. Here are the problems for Oracle (and pretty much anyone at this point):<p>1- Cross-platform ecosystem is shit when it comes to APIs (integration is normally much bigger of a challenge than the implementation itself)<p>2- Oracle isn&#x27;t good at integrating all of it&#x27;s &quot;parts&quot;. They have a strong hardware&#x2F;OS company (Sun), a strong db brand (oracle), and very strong best in breed software (Peoplesoft, etc), but they&#x27;ve been working in silos their whole lives. It isn&#x27;t their strong point to do integration (just as SAP has been abysmal at acquisitions)<p>3- The margins absolutely suck in SaaS. Look at Salesforce, Workday, Amazon, etc. They are all negative or razor thin. Oracle and SAP are used to 85% gross margin on software. This will take a lot to gut (mainly from stockholders)

The main problem that I see with this:<p>- most people dislike Oracle<p>- most people like Amazon<p>So I think at least initially, it will be about &quot;approachability&quot; rather than price and technical equivalency.<p>Everyone programmer can usually spin up an ec2 instance after about 10 minutes. With Oracle, I have the feeling that just registering to see documentation will probably be more than 10 minutes and you&#x27;d end up in a database somewhere to receive emails&#x2F;phonecalls a few weeks later.

I think new entrants to the space have no idea how hard it is to take on amazon here. Microsoft are getting a clue fairly rapidly, Google seem to have just decided to cede the space.

Remember when Ellison said he was just going to outcompete Salesforce (mid-2000s)? Interesting how that worked out:<p><a href="http://www.salesforce.com/company/oracle.jsp" rel="nofollow">http:&#x2F;&#x2F;www.salesforce.com&#x2F;company&#x2F;oracle.jsp</a>

Oracle and affordable in the same sentence?<p>Sure Ellison, sure.

I was expecting to read that a new round of lawsuits had been filed as it seems that is the only way these big companies &quot;compete&quot; these days. At least they are trying to compete on an actual product instead of just suing using a bunch of patents they bought for that purpose. Of course if competing fails, you can expect the lawsuits to follow...

Of course, Sun had its own Cloud offering pre-aquisition (SunGrid). Though it didn&#x27;t seem to be particularly successful &#x2F; popular at the time (I had a journalist reach out to me to ask if I actually knew of anyone using it at the time of the merger).<p>Many of the staff and execs left shortly after -- Lew Tucker in particular, February, 2010.

While I&#x27;m not exactly a fan of Oracle I wouldn&#x27;t dismiss them so easily.<p>Due to business reasons there are lots of applications that companies would rather not put outside their own network. Some that come to mind include HR, payroll, and billing.<p>Maybe Oracle can capture some of that value in the form of data center as a service?

It&#x27;s intriguing how often HP&#x27;s public cloud offerings are omitted from the conversation. I would think there would be a lot of interest especially because they&#x27;re a heavyweight yet are using open standards (i.e., OpenStack).<p>No disclaimer: I don&#x27;t work for HP (but I have used hpcloud)

Seems like this can only be good for consumers.

From my anecdotal experience, the majority of companies and projects using oracle products, do not actually need them, and could easily get by on free databases.

Me to OP: what&#x27;s on?