Phoronix News Bot
Oracle Solaris 11 Express Was Just Released
Phoronix: Oracle Solaris 11 Express Was Just Released
After many months of uncertainty about the future of Solaris / OpenSolaris following Oracle's acquisition of Sun Microsystems, this past August there was an announcement of Oracle effectively killing OpenSolaris while being committed to Solaris as an enterprise operating system and standing behind that, but in a much different form than what we had with Sun Microsystems. Coming out of this was Illumos coming about as an OpenSolaris fork (followed a month later by OpenIndiana) and the OpenSolaris board killing itself. Now though there's a new chapter to Solaris with the immediate release of Oracle Solaris 11 Express...
http://www.phoronix.com/vr.php?view=ODc4Nw
Junior Member
Huh?
Wtf do they mean "You may not disclose results of any benchmark test results related to the Programs without our prior consent." ?
Isn't that like against the right to freedom of speech? If I buy a car who is to stop me from telling my friends how good it is? Anybody should be able to write in the press any opinion they like about any products.
Junior Member
the difference is, you license solaris express (i.e. you don't own it) but you buy your car.Originally Posted by cheemosabe
Senior Member
Your car also has an operating system, which DOES affect how it functions. No difference here, sorry.
Junior Member
nobody claimed it doesn't affect functions. but:
1) not every car has an OS, believe it or not
2) you still license solaris and buy a car. it's just that your car manufacturer doesn't explicitly prohibit the publication of benchmarks.
if you don't understand the difference between licensing and purchasing, don't blame me.
Senior Member
And technician tells you he can see that it's misconfigured but isn't able to change it. Tool will be programmed in 3 Month .... It's one fuckin stupid bit -.-
I hate bad programmers!
Senior Member
Lots of companies have this paragraph where you are not allowed to publish benchmarks
For instance, IBM has them too. The IBM super expensive slow Mainframes have no published benchmark (against x86) because the IBM mainframe cpus are dog slow. I have written several posts on this. Google on "kebabbert IBM mainframe cpu" and you will find more articles and links on how slow the IBM mainframe cpus are
Phoronix Member
Why Oracle (and IBM) Ban Benchmarks
That is due to how Oracle (and IBM) rightly see current benchmarks as inherently skewed away from the strengths of either their hardware (IBM mainframes) or software (Solaris vs. Linux or another x86 OS); the point, however much we may hate it, is actually a valid one. Mainframes, by and large, don't compete heads-up against even eight-way SMP workstations or servers - while there are uses for one where the other is present, they normally would not face off head-to-head. The same applies to Solaris, and always has (even back in the early days of Solaris for x86, Sun did not really try to market, even for a little bit, Solaris for x86 as a general-purpose operating system - Solaris for x86 was an *adjunct* to Solaris for SPARC, and a loss-leading adjunct at that).
Multi-way x86 workstations and servers are wandering into mainframe territory because there are niche applications where sheer speed is more important than cost per solution (when it comes down to cost/TPM-C, mainframes still whip multi-way x86). Solaris shows up where the advantages of Solaris (primarily ZFS and software based on the strengths of ZFS) are more important than advantages of other operating systems (primarily Linux, but also Windows or OS X) - truth be told, the interest in porting ZFS to Linux is designed as nothing less than an attack on a niche that Solaris has held onto with tenacity - financial service back-office operations.
Going forward, it will be interesting to see how Oracle positions Solaris Express (especially the x86 variant).
Senior Member
I did not know that. Do you have links on this so I can learn more?
I mean, when IBM had the TPC-C record, the Unix server costed 33.5 million USD list price. With discounts, it went down to 16.5 million USD.
But still, 16.5 million USD for ONE Unix server is lots of money. But IBM's large mainframes are much more expensive than their Unix servers. Every calculation you do on Mainframes, it will cost you. They are brutally expensive. Mainframes are something like 0.001% of all servers shipped, but account for 10% of all money used for servers. In other words, a tiny server takes a huge piece of the cake. They are expensive, yes. And have sloooow cpu performance, any modern x86 is 5x faster for a fraction of the price.
And now you say they are actually cheap? Do you have more links on this?
No, that is not the primarily advantage of Solaris. Solaris is stable, scales far better than Linux and is cheaper than RedHat.
Phoronix Member
Where mainframes whip x86 (even multiway at present) is in volume batch transactional processing (especially realtime batch processing) - the sort of processing that requires (in terms of x86) multiblade clustering. It's niche processing, to be sure - it's also pretty much remained the strength of mainframe computing, and why there's still room for it. In terms of sheer speed, a mainframe will lose to a smaller cluster; however, as the transaction volumes increase, multiway clusters start to bog down (pretty much no matter what OS the cluster is running on). Admittedly, Solaris will bog down less than Linux (any Linux); however, even Solaris is limited to what size cluster it will support. However, Solaris is *less* limited than RedHat, and can withstand greater transactional pressures than RedHat (or any Linux, for that matter). Further, it can run on the same clusters that RedHat does (no extra engineering is required in a straight-up replacement); naturally, if the critical task is running an Oracle database, Solaris + Oracle database + ZFS is far more reliable than the same database running on even RHES due to the filesystem advantages held by ZFS over any FS supported by RHES.
x86 wins where speed is critical, and that is why low and even middle-level transactional batch processing is moving to clusters (either Linux or Solaris-based for the most part), because speed *is* becoming important. However, there are still areas where very high volumes of transactions (larger than even the burliest x86 cluster can support) are the recipe du jour; and that is where the mainframe comes in. There may come a day when the mainframe will be completely obsolete....but not yet.
Reply With Quote
