Slashdot Mirror
Oracle Linux Explored
M-Saunders writes "Two days ago Slashdot reported on Oracle's move into the enterprise Linux market, and how it may challenge Red Hat. Red Hat's stock has already dropped, and there's a great deal of talk about the implications of this act. Linux Format got hold of the 'Unbreakable' distro to find out what's going on under the hood. Is it a breakthrough for Linux in the corporate market, or just another RHEL respin? See the article for all the info and screenshots — including an 'interesting' choice of GRUB colours."
http://www.thecodingstudio.com/opensource/linux/?q =node/20
Try installing Oracle Express Edition. It installs a base database, networking and sqlplus. It even installs RMAN so you can do proper backups (the backup.sh script that comes with it is an export which is a logical copy as of a point in time for each table). You can use it for free, catch is you can only store 4GB of data. I think this was to compete with MSDE which stores max 2GB.
But my organization is not allowed to just go to any schmoe who says they support generically Enterprise Linux. There's a reason we get the contracts that we do with customers, and one of the main ones is because we use a WIDELY supported OS (Red Hat EL) that is common criteria certified to a certain level. Likewise, Red Hat has had it's certification program for professionals out for several years now, and we have several people on staff who are certified and know backwards and forwards how to install and support Red Hat as well as Oracle products.
Likewise, the licensing scheme is pretty interesting. That is NOT the price per server. That is the price per CPU...how they determine the actual CPU will probably be something stupid like their database products, where a quad-Core CPU they count as 2.5 or some nonsense.
Also, not sure how many people have called Oracle lately, but when I call for support, I don't want to be transferred to some faker in India who I can't understand, who says their name is Joe. Dell was guilty of that early on, and we saw how well that worked. Now, their Gold and Silver support for the USA is all back to 100% English speaking people usually in the CST time zone. This is the mistake Red Hat never made...when you buy premier support from them, you get access to an RHCE or higher support person in the USA who you can actually understand, who generally isn't guessing on what your problem might be.
If Oracle wants to compete with Red Hat globally (markets OUTSIDE the USA), I can see that. But I think any USA residents would be fools to go with Oracle instead of Red Hat.
Like anything Oracle tries to do after the fact and supposedly *better* than others (Oracle Collaboration Suite?), I think this idea to compete directly against Red Hat is a stupid one. When I have Oracle issues, I don't even call Oracle anymore...I call a 3rd party consultant or an engineer at Red Hat...99% of the time I usually get better/quicker results.
Yesterday I suggested Oracle entering the game would be good for the market by increasing competion among Linux vendors. Looking at this offering, I have to say: what a joke. I was completely wrong.
Oracle are pulling nothing more than a publicity stunt with this. I expect I would be correct in the speculation that some marketing executive asked some developers to slap together an “Oracle branded distribution”. They then took a release of Fedora Core and changed graphics and colors. Boom! Instant industry player.
Why bother.
That price sounds high unless you are talking the full Oracle Suite. Oracle has very good performance, is very stable, is well supported, has a clustering and failover (RAC) capabilities, built in messsaging for DB-to-DB communications, fully supports ODBC and JDBC connections, runs on almost any OS from mainframe to desktop, conforms almost 100% to the Relational DB model, supports high volume transaction rates, has row and column locking, supports encryption, can store binary large objects (BLOBs), and has a long history of success in the Enterprise. Downsides are it's hard to install correctly right out of the box, it is so flexible it is hard to "tune" for best performance, it is not something you can just "play around with" it takes some learning to handle it so good DBAs are not cheap, and it is expensive although discounts can be negotiated. YMMV...
Nope, parent is incorrect:
d f
http://www.oracle.com/technologies/linux/ubl-ds.p
That's $1199/$1999 *annually*, and "Lifetime" is defined as 5-8 years.
Take a look at this for an allegedly unbiased opinion (but who knows what is shilling and what is real these days?!).
J.
You're only jealous cos the little penguins are talking to me.
Back in 02-03 I worked for a small startup. We were running Oracle on Linux doing dev work. We called them up to inquire about licenses. I think we were quoted $32k for our setup. We naturally told them, nevermind, we'll port it to MySQL and they eventually came back and offered us a deal at $4k. Of course, our app was meant to be installed at several high profile insurance companies so that meant more Oracle Licenses for them in the future.
BTW, all those numbers are from my rather fragile memory. YMMV.
-- Jason
Remember that we are talking list price for one server.
I can speak from experience that Oracle's architecture is better than DB2, substantially better than SQL Server, and completely blows Sybase out of the water. Oracle 7 or 8 years ago was handling concurrency and large transactions better than Sybase does today. The CBO is much better than everyone's except maybe DB2. The hardware support is broader than just about everyone else with the exception of DB2. Locking is better handled. Indexes are efficient even on columns that aren't integers. VARCHAR support is clean. PL/SQL is quirky but less quirky than the alternatives. The trigger support is richer.
What generally happens is that a customer will go with Oracle for a handful of critical apps that justify the high price. Then once Oracle has their foot in the door, they'll come back and offer an expanded deal to host the databases that could run perfectly fine in any db, and do it all at a discount. The end cost is going to be substantially less than one would suppose by scaling up the quoted numbers.