Mike Olson on Oracle’s ‘Unbreakable Linux’

It’s no secret that I’m not a fan of Oracle’s attempt to co-opt Red Hat Enterprise Linux. But I have to admire Mike’s recent post on the topic. Mike makes no apologies and lays out very good, business-driven reasons for Oracle’s use of RHEL. Again, I still believe that just because you can do something doesn’t mean you should. But let’s listen to Mike:

Red Hat Enterprise Linux has established itself as the most popular distribution for enterprise use. When Oracle considered the best way to add Linux support to our services portfolio, we considered our customer base — primarily enterprises — and made the obvious choice.

As a result, Oracle is able to offer services to enterprise users who already have RHEL installed. This is crucial. The leading role of the RHEL distribution in the enterprise is exactly why Oracle chose to support it, and not some other distribution, in our Unbreakable Linux support offering. Enterprises running RHEL today can switch to Oracle’s services without downloading a new copy of the operating system. Naturally, any enterprise running production systems is reluctant to reinstall anything at all in order to switch vendors. It’s just too risky to change the tires while the car is moving.

The effect is that Oracle is able to convert downloads to paying customers without requiring a download.

An interesting and valid point, and goes a long way toward explaining why Oracle’s download numbers have been so low. It’s parasitic, which I don’t like. But it has been effective, and it’s the customer’s voice that matters most.

But here’s what I don’t like. Oracle’s Linux depends on Red Hat being successful without actually giving money to Red Hat to help it be successful. Quite the opposite – it actually sucks money out of Red Hat, potentially undermining the very ecosystem it’s building its Linux support on. I guess the rationale is that if Oracle’s Linux does succeed in killing Red Hat Enterprise Linux, by that time Oracle will have enough momentum that it could take over as the lead Linux distribution.

But this is precisely what I don’t like about the Oracle move, Mike. It doesn’t build the Linux ecosystem. It draws from it, without giving back. Yes, I know that Oracle contributes to the Linux kernel. That is not in dispute. Oracle has been doing this for a long time. (I remember Rene B. speaking at the Enterprise Linux Symposium back in 2002 and talking about its contributions to Linux, though he made it clear then as to why: cheaper operating system, commodity hardware, and Oracle databases. Everything but the database gets cheaper, making customers think the database TCO is lower….Shrewd.)

But the argument you’ve laid out depends on Red Hat’s success, without contributing to it. I can see a range of ways that Oracle could have done this – could still do this – and get the “superior support” that Oracle has suggested is the reason behind the move. If this is so, why not work with Red Hat rather than against Red Hat?

You correctly identify that service/support is the wellspring for Oracle’s Linux success:

Fundamentally, though, [our] success is driven by the work of the team that produces and supports [Unbreakable Linux]. Oracle’s long-term success with Linux depends on our continued collaboration with the Linux community, and on our ongoing contribution to the kernel and related software. It’s only by investing in that work that we will continue to succeed commercially.

But what you say before that is that your success depends upon building on the success of another company, which success you are no longer contributing to in the same way that you did as a Red Hat partner.

Your post makes a lot of things much clearer, and I’m willing to give you the benefit of the doubt that the success has been as real as the customer testimonials point to. But don’t you think that your success could be that much more real if you were working with Red Hat on this, and not against Red Hat? I mean this sincerely.