Oracle to Amazon: It’s on
gigaom.com
gigaom.com
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/tools to the red tape.
The bug tracking system looks like an HTML 1.0 web page. The "database version" is a required field, even though the product you work on doesn'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't even bookmark things.
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't even take into account the legal approval steps that need to happen which also takes a while.
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't even try and use 3rd party libraries because it'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 "low cost center", the library ends up sucking.
Recap: Amazon has nothing to worry about, because nothing innovative or powerful is going to come out of Oracle.
But, you might have basic misunderstanding of Oracle as a whole and position in the market. Oracle is an excellent acquisition and integration machine: They did excellent job with BEA, PeopleSoft, Sibel, GoldgenGate, etc. That is the key of Oracle: very well tuned system (marketing, support, sales, engineering, DDR, etc.) which can make a lot of money in enterprise environment.
From employees perspective:
At least in my org, integration is still sputtering along almost 2 years after I have left. Oracle bought another company shortly after us that we had to integrate with, in addition to Oracle proper. Over the last 2 years, most of the more competent engineers have left because they couldn't take the work or environment anymore. The end result was more work for everyone else and inability to find new employees due to the hot market. New feature development are practically non-existent.
On the non-technical side, there was nothing excellent about integration. We lost vacation time/holidays. We didn't have wireless for probably 6 months at the office because our wireless wasn't "secure enough". Everything worked less well than it did before, from phones to the network.
From the customer perspective:
For the customers that don't care about whatever product you are integrating with, they haven't seen many new features because everyone is focused on integrations. There are now less people that understand the product focused on you.
For the customers that do care about the integrations, at least in my org, the integrations have been so shaky I can't help but think there will be some buyers remorse. And again the quality of people supporting you is now lower.
I'm skeptical Amazon has anything to worry about.
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 'added value option' and high touch management / maintenance. Not something cloud compute is about.
To my way of thinking it seems like step one here, would seem be build a PUE < 1.05 data center with 10MW of provisioned power. Have they done that?
They have the financial resources of Google, dwarfing Amazon in that regard. I don't think any of this will be decided by who has large data centers - that's a commodity that can be acquired at will when you're talking about companies with tens of billions in the bank. This will all come down to execution.
Amazon gives you access essentially to mostly open source tools or licensing fees for closed source stuff. it seems almost a different market to me.
Iduno, the hosting aspect of the Java world is so weird. My Python/RoR friends try to tell me there is no good Java hosting & then I point out the fact that its right there on the Heroku they love so much. I think cloud-enabling Oracle will just make the massive power of the JVM more apparent with better tools (yes they may cost some money but so do Heroku add-ons)
That said, I do think the Java community should continue developing it's "rapid dev" sortof languages/frameworks (Clojure and maybe Scala/Play?) if they are ever going to attract start-up and small business types.
But yeah Java gets such a bad rap... I've had way less headaches since moving over from RoR. Java is a little more DIY in terms of composing a toolset (if you happen to be starting from scratch -- many like myself are not) but it's all there & way powerful. I deploy a handful of pretty well encapsulated projects to Websphere Application Server and it's pretty much a breeze, plus the IBM products, for example (not Oracle I know...) have massive admin interfaces that a rapid dev team couldn't mimic/maintain even if they wanted to.
The open source tools are great too, but as many like to complain -- Spring and the like "are just overkill for small projects". That may be true but why even put all this thought into your toolset if you are just trying to make small low-traffic CRUD sites...? Iduno, strange world.
And Oracle does dwarf Amazon, significantly, in terms of financial resources. That's not a subjective opinion, it's an easy direct comparison.
Amazon has $7.6 billion in cash, a pile that is contracting rather than expanding, and they have to fund the continued expansion of their retail business, investing billions into new fulfillment and automation. Meanwhile it's highly unlikely Amazon is earning much profit out of AWS itself. Amazon produces some cash flow, but they do not produce net income, and do not compile much cash to their balance sheet from operations.
Oracle has $39 billion in cash, net income of $11 billion, and even more cash flow. Out of $37 billion in revenue, Oracle generates $14.6b in operating income; Amazon generates approximately zero operating income.
To put it all in perspective, Oracle generates more net income per quarter than Amazon has in its entire history. Oracle generates more cash per year to its balance sheet than Amazon has in the last ten years.
Saying they are comparable is saying that when Oracle sells a DB license and Amazon sells a toaster, these are only different in the amount of money the generate. They both have equal experience with software, even though Amazon's experience involves putting an item in a box and putting a label on it.
0. http://www.informationweek.com/cloud/infrastructure-as-a-ser...?
And you're obviously including something in calculating Oracle's size, since Oracle's cloud business is $0 IIRC -- which is much smaller compared to AWS's $3B.
Try $1 billion annually in saas revenue
They can't afford to cannibalize their existing cash cow.
That's still a very lucrative market, but I would guess that options from both the low and high end are starting to erode new sales opportunities and they might be losing some outright. e.g. it used to be the default in government IT systems to just stamp in "Oracle dB" in high level architecture diagrams and not get fired, but now postgres and other alternatives are being seriously considered and used.
"Oracle dBA" is being replaced with "dBA with experience in X, Y and Z databases" in many job reqs as well.
They still have huge enterprise license agreements, particularly in the fed space, but as budgets get cut those are simply being reconsidered.
This isn't really a move for "today", but for where Oracle will continue to vector if nothing is done over the next 3-5 years.
I call bullshit. No one is making "Oracle-like margins on enterprise applications" in the cloud...yet...not anytime soon in this competitive market.
This is the opportunity in IaaS/PaaS as I see it: every large company in the world typically outsources their operations systems infrastructure. Here's an example - a company based in Europe that does €7b in revenue. Let'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'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 "Alright we need 6 HP ProLiant servers, Oracle 10g installed on 3 of them, and SAP HCM installed". This process takes months. To everyone in the web world who'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're "professional" and "enterprise grade". So here's the opportunity - do what Amazon did for the web in the enterprise. So, let's say you want SAP HCM, well that'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):
1- Cross-platform ecosystem is shit when it comes to APIs (integration is normally much bigger of a challenge than the implementation itself)
2- Oracle isn't good at integrating all of it's "parts". They have a strong hardware/OS company (Sun), a strong db brand (oracle), and very strong best in breed software (Peoplesoft, etc), but they've been working in silos their whole lives. It isn't their strong point to do integration (just as SAP has been abysmal at acquisitions)
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)
Workday has a profit margin of 62% versus Oracle's profit margin of 82%. It's less but it's not that big of a gap.
- most people dislike Oracle
- most people like Amazon
So I think at least initially, it will be about "approachability" rather than price and technical equivalency.
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'd end up in a database somewhere to receive emails/phonecalls a few weeks later.
But regardless of my feelings, only the people making the decisions on where to spend the money matter. Oracle has made it their business to be very good at converting those people.
My take on it was that Google's pride in their own engineering would not let them do that. They seem to be in it for the long haul here.
Sure Ellison, sure.
I've had better support experiences when I can dig into the source, say linux or a kernel module, or a driver and figure things out on my own while I wait to get punched through various level 1/2/3/etc.. support people. I vastly prefer open source for this reason alone, with oracle stuff, its basically a black box that barely works if you look at it wrong. This is with everything fully automated mind you, I still see jenkins test failures where installs will not work seemingly at random. I can debug or retry the job and it'll work. Its my most hated piece of software right now.
Many of the staff and execs left shortly after -- Lew Tucker in particular, February, 2010.
Amazon hit at the right time: virtualization as a concept was being adopted by developers, storage needs were going through the roof, classic storage systems weren't scaling at reasonable costs.
Amazon's other strength is that they can use the massive revenue they have sitting around for short periods of time to build out their computational infrastructure (turning their operational revenue into capital spending). IaaS took off; SaaS somewhat, and PaaS almost not at all (except in a few well defined areas).
I can't really see what oracle woudl bring to the game. It looks to me like they are just making a play to take their existing customers and move them to their data centers.
Sun faced multiple challenges, largely being a legacy product with legacy offerings in a world where price point was key, their service commanded a premium, and much existing infrastructure wasn't positioned to properly capitalize on it. That's just off the top of my head, though I'd be interested in seeing what other factors accounted for the failure.
Oracle has the kiss of death for far too much in my book. I'd consider any offering it presented with extreme reluctance.
The former certainly suffered from platform constraints and a fairly rigid 'stack', but the latter was far more flexible and various teams within Sun worked closely with several potential customers on the proof of concept. Unfortunately this fell apart thanks mainly due to the commercials and some concerns with how the platform was configured within the context of security and ensuring appropriate segregation of data.
I like that the last sentence you wrote could easily describe today; many customers still care very much about the security context and the data segregation.
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.
Maybe Oracle can capture some of that value in the form of data center as a service?
No disclaimer: I don't work for HP (but I have used hpcloud)
"This phrase is used after a bet has been made."