Confrontation between HP and Oracle. Continuation

    After reading the note “HP and Oracle customers react: not happy, but coping” on the Forrester Research blog on zdnet, I decided to translate it and add some comments to my last post “How will the opposition between HP and Oracle affect Open Source?” . Translation under the cut, but for now a few comments on the last post:

    Despite the fact that someone was struck by the “analytics” of the document, in fact, Oracle’s rejection of Itanium is predicted there. Which, in my opinion, testifies to the professionalism of the staff who wrote "A Brief Analysis ..". Perhaps, thanks to that document, the company saved a lot of money.

    Regarding comments of the form: "And what does OpenSource have to do with it, I still don’t understand ...". I want to draw your attention to the fact that a Forrester post says that most customers consider the possibility of switching to Linux (as in the case of “Brief Analysis ..") as the most likely way out of this situation. And on zdnet, a post about the reaction of HP and Oracle clients is in the Open Source section .

    In the comments to the last post, it was not in vain that IBM's DB2 was mentioned (tackle details) - currently only IBM and EnterpriseDB have transition mechanisms from Oracle . By the way, after reporting Oracle's abandonment of Itanium, EnterpriseDB decided to provide support for HP-UX .

    Further retelling of Richard Fichera's post:

    HP and Oracle customers are not thrilled.

    After Oracle hit HP and Itanium, I sent out a lot of emails to various companies and received about a dozen messages from HP and Oracle customers who wanted to discuss further options. I did not notice any kind of mass panic in the answers, and I summarized the scenarios of the development of events into several conclusions:

    Most clients of HP and Oracle have not yet switched to the latest version of the DBMS, so they have time to take any measures. This time is limited by the need to update their applications and DBMS to a version larger than the current Oracle DB 11.2 for Itanium. It turns out that for those clients who are still on the 9th version of Oracle, the time for making decisions can reach several years. But those who are already at 11.2 today will be forced to migrate to something along with the next update cycle. The most common application that was mentioned in the messages was SAP, the next most popular were the programs from Oracle itself.

    Customers with Oracle software such as Hyperion, PeopleSoft, eBusiness Suite, or from third-party vendors are likely to face severe restrictions on updating these products. In some cases, the decisions of independent developers will stimulate users to the changes that they do not really want. Several clients told me that they would postpone updates to avoid the need to use unsupported versions of the DBMS.

    So what are the companies described above going to do regarding their server platform? Most customers do not plan to change anything, at least in the short term. They want to consider all possible options and deliberately choose an alternative. I would say that only two clients have a “panic mode”, and in both cases this is due to the previously existing plan for a large-scale upgrade of HP servers and Oracle software. These two companies, frankly, need to find an alternative over the next 6-12 months.

    However, the absence of panic does not mean the absence of anxiety. All current HP customers are considering options that come down to one of three (or can be done simultaneously):
    - Let it be as it is. Most of the clients I spoke with are going to wait a while to see how realistically the appearance (over several years) of Superdome class systems on x86 and HP-UX. My personal opinion on this is as follows: HP has the technology to do this, but there is no guarantee that Oracle will support HP-UX on x86. I would advise against this.
    - Switch to Linux. Customers who had plans for migrating to Linux told me that they would be more active in implementing these plans. Considering the Intel Xeon E7 lineup and further improvement of Linux, most customers will consider this option. However, there are doubts about Oracle's readiness for long-term support for RHEL or SUSE. Several customers have reported that it is rumored that Oracle will only support its own Linux distribution in the future. It goes without saying that we will turn to Oracle for clarification.
    - Switch to another UNIX provider. Two companies reported that their only reasonable option is to switch to another UNIX system that will support future versions of the Oracle DBMS, as well as applications from its stack. For most modern enterprises, this means choosing either IBM or Oracle / Sun. Ironically, one of these two companies has just completed the multi-year transition process from SPARC / Solaris to Itanium / HP-UX.

    Customers are extremely concerned about this situation. Everyone with whom I spoke feels that they have received a “knife in the back” from Oracle and the general idea that runs through all the discussions is “How can I trust Oracle if they treat their customers like that?” The reality is that most customers have virtually no choice - they should either stay with HP-UX, freeze existing solutions as much as possible, or switch to x86 and Linux. In addition, Oracle's abandonment of Itanium has strengthened the desire of some customers to switch to Windows / SQL. And a whole group of users has formed that would be grateful for any attempts by HP and SAP to provide their "Oracle-less" solution.

    Of the competing UNIX vendors, only IBM can turn this situation to its advantage. Oracle software does not work poorly on IBM POWER servers, and IBM has a viable offer for migrating from Oracle to DB2. There is a risk that Oracle will try to repeat the focus that they did with Itanium and with IBM POWER (after all, any sales logic applied to Itanium compared to x86 also applies to POWER), but I think that this will not happen soon. For various reasons, such as fearing customer anger, panic, or retaliation from IBM, which is not as defenseless as offering its own DB2 database engine.

    Also popular now: