Oracle Bullies Enterprise Clients Into Cloud Purchases, Consultant Claims

An anonymous reader writes: A consultant claims that Oracle has adopted the widespread use of 'breach notices' this year to force existing enterprise customers to adopt its newly-bolstered range of cloud services, or else be told to stop using all Oracle software within thirty days. Speaking to Business Insider, the unnamed source described the tactic as a 'nuclear option' which is now practically the default when the need to add services or users to an existing contract triggers an 'audit' by Oracle. An ex-Oracle contract negotiator who now works in the ever-expanding business niche of 'Oracle contract negotiation' commented 'Internally, the water cooler gossip there is that they've never seen this kind of aggression before. Oracle has really dialed it up. Customers are buying cloud services to make the Oracle issue go away, not because they have any intention of using cloud services.'

Comments

How much you got?

[pigiron]

After dealing with Oracle for over thirty years I've learned that the answer to the question "how much does Oracle cost?" is "how much money do you have?"

Re:How much you got?

[Dutch Gun]

Seems like a desperation move for a company with under-target earnings, if they're willing to poison long-term relationships with their customers like that. You're going to see businesses deciding that they don't like having a gun held to their head. They'll pay the ransom for now, but some of them will probably start investigating other options in the background.

Re: How much you got?

[Anonymous Coward]

Oracle is not particularly buggy. It's a stable and mature piece of software. However, it's extremely limited, always has been. Like Cobol, everything is limited so it puts real constraints on the solutions you build with it.

Choosing Oracle is not really about support or quality though, it's about CYA. See, the people who makes the decisions don't really know what they're doing, but nobody was fired for choosing IBM^H^H^HOracle, so it must be a good fit for your system, right?

Wrong. There are many alternatives and better tools for various problems: https://duckduckgo.com/?q=oracle+alternatives
Even some who provide enterprise support.

Anyways, SQL is really really old now, and is only chosen because people generally don't like to think for themselves and open themselves up for risk. Well, then you're also becoming a dinosaur who uses IT as a cost center instead of advancing your business with evolution of technology. As soon as top management openly admits that, everything is good, because then they're finally being honest at least. It won't help the business though until they become expert at managing IT services, which today is part of every fucking company.

Re:How much you got?

[Anonymous Coward]

This is just what happened at my previous job. The ever increasing cost and pure hostility against customers made the company start switching from Oracle to other alternatives.

Re: How much you got?

[cyber-vandal]

MongoDB is webscale!

Re: How much you got?

[__aajwxe560]

I'll go a step further - a few years ago, was looking to add SSO to web product for our customers, and as an existing Oracle customer, they had a package offering for SSO from acquisitions. After agreeing to pay for some base consulting to get preview going, quickly realized what a pile of overcomplicated shit it was at the time (the consultants flown in from all over the place couldn't get the base "hello world" even working, and they all claimed to do this all day long and not unusual challenge). I saw if they couldn't get going, didn't want the ongoing support nightmare, pulled the plug and went with something much simpler. Oracle sales guy laughed, said he already added the skus to our upcoming databse support renewal, and magically, if I tried to take them off, we would loose "bundle" pricing and costs would go way up for just our existing db licensing (ignoring the fact that they were effectively charging us 30% more regardless for no new actual software). One example of many. Oracle is the one company that can fuck off harder than Microsoft for their shitty biz tactics. They shoot themselves in the foot in big picture - any new companies u am at, I steer to anything but Oracle.

Re: How much you got?

[gweihir]

MariaDB is only good if you need extremely fast. For a full ACID compliant DB, go for Postgres.

Re: How much you got?

[Nkwe]

Do you think gmail would have become the most popular email service if it would have used ACID? ACID is *not* the only option. It's the *old* option. It's the expensive and slow option.

Maybe ACID doesn't mean what you think it means. It is not a technology that is "old" or "new", it is a way to think about the requirements of your system. Each of the four letters in ACID stands for a particular property of a database system and these properties (in various combinations) may or may not be needed by the system being built.

If your system is processing something where the integrity of the data is important (like financial systems), you are very likely going to need all four properties. If you are moving money from one place to another, you want to guarantee that that the the money is completely moved or not. You don't want the money partially moved, you don't want money to be lost, and you don't want money to be created out of thin air. ACID (as a concept) guarantees this.

If your solution requires ACID, you don't have to use a database that supports all of the properties of ACID, you could instead implement ACID in your application layer. However if you do this, you have to guarantee that that your application layer implements it properly and that there is no possible way to get to the underlying data store without going through the application layer. You also have to guarantee that no changes, updates, upgrades, or bugs in your application layer every break the ACID guarantee at any time. Making all of these guarantees in your application layer is VERY HARD, which is why people use ACID complaint databases instead to solve this particular problem set.

If your requirements don't need the properties described by ACID, than there isn't anything wrong with using a non ACID database. If may be acceptable for your data to "eventually" become consistent, to be inconsistent, or maybe even lost.

In the gmail example, you don't really need all the ACID properties, so you don't need to use that sort of database to hold the information. Email is not transactional end to end; when you send an email you are not guaranteed that it will get there. Email is also not order guaranteed; if you send multiple emails there is not a guarantee (or need) for them to arrive in the destination mailbox in order. If you are bulk moving messages from one mailbox to another, and only some of them get moved, it is okay and you can just move the remaining messages later.

As always, it is important to chose the right technology to solve the problem you need to solve. ACID compliant databases solve a lot of important problems (usually involving money), and if you have one of those problems, there is nothing "old" about ACID.

Re: How much you got?

[St.Creed]

Actually, the last time this was debated here I brought up a similar point, and someone else pointed out that banks don't use ACID but mostly use eventual consistency for their transaction systems. That does cause them to lose (a lot of) money sometimes, but they write it off against the expenses of real-time ACID compliance.

Headline from the future:

[Gravis Zero]

The End of Oracle: Unhappy Customers Jumping Ship In Droves

You can only be pushy for as long as you are irreplaceable.

ORACLE is

[wbr1]

One Raging Asshole Called Larry Ellison

Danegeld

[MikeRT]

This is nothing more than corporate Danegeld. It will probably end about as well for most of their customers. They'll be just shocked when Oracle comes back in a few years and launches another attack on them.

Virtually every database I've ever seen is a bit bucket. There's precisely zero reasons for them to be on Oracle because the data set is well into the size where PostgreSQL, MySQL and SQL Server could easily provide a more cost-effective alternative. If you use Oracle for that, you get what you deserve.

Oracle is just doing this crap because they've realized that nobody really wants them for big data. They know that their future is mainly limited to the sort of customers that are willing to buy and build SQL databases for their data. There's plenty of legitimate room for that sort of data and they'll do fine. They just can't accept that they're on the infrastructure side of cloud computing and big data that corresponds to where Microsoft is in mobile.

Heck, Microsoft at this point should black knight them by releasing a trojan that infects company networks and all it does is audit their Oracle stack and send Oracle sales an email telling what it finds on the company network.

Re:We're ditching Oracle

[Shados]

And this is where things are getting dicey for Oracle. Even an open source DB can deal with billions of rows. And when you go beyond that, people start using multiple interconnected specialized systems instead: a big mismatch of a relational db, hadoop, redshift, dynamo, vertica, spark, etc.

If you need a trillion records in one table, there's better commercial options than Oracle. If you can need specialized tool to handle different data sets of various size, you'll be using a soup of tools, most of which are open source.

There's no reason to use Oracle stuff anymore, aside for legacy compatibility, or if you use their ERP (which for large Retail, is probably the best one, unfortunately)

Re:What does Oracle do well?

[pigiron]

High volume ACID transactions and stored procedures.

Workday learned their lesson...

[erp_consultant]

All of the Workday executives are former PeopleSoft executives. PeopleSoft (now owned by Oracle after a nasty takeover battle) is a great product but it has a fatal flaw - nearly all of the critical components are controlled by someone else. Database (either Oracle, SQLServer or DB2) is owned by someone else. Middleware (WebLogic) is owned by someone else. Reporting (SQR and Crystal Reports) is owned by someone else. Hardware is owned by someone else. Operating systems are owned by someone else.

Workday, starting with a clean slate, decided that they wanted to control everything. So they used an object oriented open source database. They own and control every layer of the software stack. They, since it is cloud based, control the hardware.

This gives Workday a big advantage when it comes to supporting the software. There is only one configuration to support. Oracle and SAP and others have hundreds of combinations of database, hardware, operating system, etc. to support.

Oracle has typically been able to use its stranglehold on the database platform to force customers do this or that. But they can't do this to Workday or its customers. And this has Oracle scared shitless.

Oracle is rushing to get cloud based products to the market. I don't know that Oracle is trying to strong arm their customers into using those new products but it is not without precedent.

What I do know is that internally they have this philosophy known as TOTO (Turn Off The Oxygen). That is how they destroy their competition. Their hope is to TOTO on Workday until they run out of money and fold. They know that Workday is operating at a loss and that their stock is trading at insane P/E levels (2650 as of Fridays close). Oracle will give away their cloud offerings if they have to. It's a waiting game and Oracle has the cash to wait it out.

Re:Your data held hostage in Oracle cloud

[St.Creed]

I'm using SQL Server 2012 now (first time I ever used SQL Server for serious data loads) and I have to say it performs pretty good as a data warehouse for a moderate size organisation. We're loading 500 million lines and it seems to hold up well on a single mid-range server. Querying the whole set is not a pleasant experience if you do a full scan, but if the index is selective enough we get okay performance out of that as well.

Over the last decade most of my deployments were on Oracle but I think that for almost any business I know, SQL Server is a pretty good alternative. I'm not so impressed with the query performance but update/insert performance is much better that I know of Oracle.

However... if you need decent materialized views, or analytical functions, or really low-level control over the database, Oracle is still the first contender. Statistics are easier to manage on SQL Server, though.

However... the SQL Server pricing is not as low as it once was, and climbing steadily into Oracle territory. So unless Microsoft can keep the price down, it may not offer much of an alternative.