Performance advantage of RHEL

(From another thread)

Thanks. From the article:

1. What are these improvements in kernel, virtualization, security, power saving etc?

2. Are they available only to RH or has RH contributed them to the general community?

3. What is RAS?

Some big questions there 'borg.

1 - Try to get hold of Red Hat's core kernel devs and ask them. Then report back here.

2 - I assume they all go upstream. GPL and all that..

3 - Reliability, Availability, and Serviceability. Big acronym in datacentres.

This post has been edited by spoovy: Jul 11 2011, 07:09 PM

Wha...RH doesnt really technically say what the improvments are, and rather makes politician like generic statements in their media releases?

Read any technology company's promotional bumf and you will see exactly the same thing. Do you really expect hugely technical information in press releases? And would you even understand it if you saw it?

If you genuinely do want information on a specific topic, such as power saving improvments, then you need to do some research - contacting Red Hat, as I suggested, would be a logical first step. Please do, then report back here as there may be others on SLF who are interested.

This post has been edited by spoovy: Jul 13 2011, 07:50 PM

Wha...RH doesnt really technically say what the improvments are, and rather makes politician like generic statements in their media releases?

Do you find that strange? Why do you think they cover up updates to their kernel or any technical info?
Not to confuse the average user, but to block things Oracle is doing (copying their stuff and selling support for the work Red Hat did).

So yeah, press releases are always misty.
Just enough info to get people interested but not always the full details.

There are lies, damned lies, and statistics. ...and then are press releases. It's been a few years since I had to do public relations, publicity, marketing, and advertising. I still feel dirty...

But I digress...

RH could, conceivably, identify specific performance improvements. However, considering the number of packages included in RHEL, and considering that one man's improvement is another man's disenfrancisement (KDE 4, GNOME 3, virtualization supersturation, and GPL ultra-orthodoxy spring to mind), actually documenting such improvements would be burdensome, futile (most programmers toot their own horns, anyway), and subject to the same flaw that all benchmarks suffer from: no common point of reference, and meaningless drivel to the ordinary user.

Well... there is WUProp@Home, but it's neither not meaningful to the general public nor remotely useful for comparing OSs or architectures outside of BOINC.

With respect to security... aye, there's the rub: you can't sell absolute security. Even a little puffery on this subject could land you in court, as Apple recently discovered with their "What Macs viruses?" line of advertising. And as for power-saving, that's predominately a hardware issue.

It's also a budget issue. How much would it cost you to change all incandescent lightbulbs to compact fluorescents? Or LEDs? Now try and convince middle-management that you need to replace all your hardware because the latest models consume only 5% less wattage at a cost of at least double next year's budget and still would have to be replaced in 18 months' time just to keep up with the latest models competition's latest-and-greatest...

* sigh *

The nice thing about RH is that it does release its code into/onto/unto the community just as much as RH borrows/incorporates/gloms from the community - both as source code and in binary form (I know you're listening, Cupertino!!! ). In this sense, RH's advantage is that it has taken full advantage of "crowdsourcing" (PR words... yuck!) and the near-instantaneity of the Internet to maximize the quality of the product they produce (or have produced for RH) and curtail any damage "in the wild" by having security updates, bug fixes, and patches available as soon as they're ready. In this near-real-world sense, RH (and its progeny) has a massive advantage - a definite "improvement" for anyone migrating over from [the leading] proprietary OSs.

However, since "crowdsourcing" (Anyone got a barf bag?) is now The New Black Normal, you can't put it in a press release. It'd be like insert CPU manufacturer here making a big deal about the FPU, MMU, ALU, and NYU in their latest chips.

Yes, "NYU" was bubkis - just like in real life

Wha...RH doesnt really technically say what the improvments are, and rather makes politician like generic statements in their media releases?

Do you find that strange? Why do you think they cover up updates to their kernel or any technical info?
Not to confuse the average user, but to block things Oracle is doing (copying their stuff and selling support for the work Red Hat did).

So yeah, press releases are always misty.
Just enough info to get people interested but not always the full details.

Not to disagree with you or anything, but that's not a totally fair statement about Oracle's Linux Distro, as that same statement could be claimed about all the open source distro's based of Red Hat.
Oracle like the others (Scientific & CentOS etc) even state that their disto is a downstream version of Red Hat. The only difference is Oracle is more about profit margins (i.e. support & software) & attempts to improve upon red hat with their Unbreakable Linux Kernel, and even then you can freely get their linux distro with the Unbreakable Kernel . Which you can make a pretty good OS from it, without even needing to pay for support (Through there new Public Yum repo).

I am currently running Scientific & Oracle Linux with just open source repositories, and so far the only difference i have found is that it requires a little more work to set-up Oracle with repos than any of the others.

If you were to criticize them for anything it would be turning Open Solaris into "Closed Solaris".

You can get the distro for free, but have to pay for security updates and other patches or enhancements.
Check their info on the Oracle Public Yum Server:

In my opinion it is just a commercial rip-off from RHEL. And that is something that I find Oracle unworthy because they are a large company with enormous resources (manpower and finance). Of course they needed their own distro since they sell a complete package to their customers (knowledge and technology) but this could have be done in a more professional manner.

It's a weird form of parasitism where the parasite would probably quite like to kill its host. I wouldn't trust the company with anything in my data centre (if I got to make those decisions; which I don't ) For me any company has to meet a basic ethical standard or I don't trust it.

I agree with you that for such a big company they do just rip-off RHEL. I suppose what it really comes down to is the fact that RHEL provides their source code openly to the community at the same time as gaining funding for that source code through its support & knowledge base of commercial businesses. Where as Oracle is more like trying to make a rock bleed, not openly giving even older source code back to the community without a asking for a fee.

I give them credit for improving RHELs Kernel with there Unbreakable Kernel which in all honesty is stronger & faster, but at the same time there is no reason why they can't provide the older source code of the Unbreakable Kernel, so that Linux in general could improve.

@spoovy: Your right spoovy, they probably would, but i don't think it would ever happen. When you look at it Fedora is Red Hats Upstream version which is pretty much 100% community based, so even if Oracle was to knock-off Red Hat, you would most likely find that Fedora would branch off and create an Enterprise Version of itself (re-invent the wheel so to speak) & it would start again.

That's a good way of putting it actually. Oracle's top man and CEO is one of the wealthiest men in the world now. One would think he could give back a little