UNIX Consulting and Expertise
Golden Apple Enterprises Ltd. » Posts for tag 'SUN'

Comparative benchmarks: here there be dragons Comments Off on Comparative benchmarks: here there be dragons

Part of the work I’m doing for a client at the moment involves a migration of databases from a Sun E2900 running Solaris 8 to a cluster of M3000s running the latest build of Solaris 10. This is going to be a long post, but I want to touch on some of the difficulties businesses face when trying to benchmark such migrations.

Initially the migration was sized on the simple (and flawed) premise that we have 24 cores running at 1.9Ghz (US IV+), and we’re now moving to 8 cores running at 2.9Ghz (SPARC64 VII). Therefore the box will be X times faster, where X is a nice big number that’s acceptable to the business. Project signed off, cheque printed – next!

This is great, but from a technical point of view it’s dangerously misleading – and it also leads to some pain when trying to manage this infrastructure. Could we consolidate more databases onto this machine? What’s the performance implication? What’s the real performance gain we can expect from the migration, and what are the key factors that determine this?

First of all we have the OS. Solaris 8 is the Honda Civic of operating systems. It’s been around for ages, it’s reliable, it’s not fast and flashy, and you can try pimping it out but you’ll end up wasting money and looking foolish.

Solaris 10 has so many improvements in areas that boost performance (as well as decreases the administration pain) that they might as well be different OSs. The only way to do a proper comparison is to sit down with the same hardware, and the same application, and run through some performance tests with both 8 and 10, then compare the results.

The problems for all businesses is that this just isn’t possible – this is a stable production machine. No-one is going to buy another one when they’re going to migrate to new tin, and when the machine becomes freed up it will be post-migration – which would be too late.

Then you’ve got the actual hardware. The UltraSPARC IV was a cracking bit of kit, but it’s been out for almost a decade now, and Fujitsu’s SPARC64 is faster in every area. There’s a huge technology gap, and again, unless we get the chance to compare the same OS with the same application on the two different sets of CPUs – but then the server hardware is so radically different that the results will be out of whack.

Let’s add into the mix some performance improvements within the M3000 itself to highlight this. Initially the SPARC64 VII CPU that was offered has a clockspeed of 2.52Ghz, but the kit for this particular project has 2.75Ghz CPUs. That’s only a 9% performance increase, right?

*bzzzrt* Computer says “No”:

  • faster System clock at 306MHz
  • faster memory – now using 667 MHz DIMMS (the memory runs at 612 Mhz, doubling the system clock)
  • faster interconnect to the Jupiter System Controller at 1224MHz (four times the system clock)

This actually adds up to a 23% speed increase for the CPU overall. This highlights the effectiveness of balanced RISC computing platforms, compared to the single minded “clock speed is king” focus of x86. It doesn’t half make of a mess of your benchmarking figures, though.

Mr. Benchmark has a great post that goes into far more depth about this sort of thing, and I highly recommend reading his blog.

The conclusion? Simple “compare speed and number of cores” comparisons are very, very dangerous, and can lead to a massive over-spec of new systems. However, many times this will be the only way to make such a comparison, unless you can get early access to a vendor’s test lab. In short, tread carefully, and think carefully about *all* aspects of system performance.

Oracle Premier Support on Dell and HP – why this matters Comments Off on Oracle Premier Support on Dell and HP – why this matters

So there were a rash of news reports about Oracle pulling third party support contracts for Solaris x86 – specifically with HP and IBM. It seems this was all a storm in a teacup, handled with the usual communicative excellence from Oracle (seriously, guys, *come on*).

Oracle have announced new support agreements with HP and Dell. This is important for a couple of reasons, which I’ll explain here as the mainstream press seems to have conveniently glossed them over in their rush to backpeddle on their previous “OMG Solaris is dying!” claims.

Firstly, it’s becoming clear that Oracle are having a lot more success than Sun when it comes to showing other x86 server suppliers who wears the trousers. Sun’s line was pretty much “Your customers want it, so go on, sell Solaris licenses and support – please”. Oracle’s new deal with HP and Dell is a lot more impressive:

Oracle today announced Dell and HP will certify and resell Oracle Solaris, Oracle Enterprise Linux and Oracle VM on their respective x86 platforms.

and:

Customers will have full access to Oracle’s Premier Support for Oracle Solaris, Oracle Enterprise Linux and Oracle VM running on Dell and HP servers. This will enable fast and accurate issue resolution and reduced risk in a company’s operating environment

It’s not just Solaris – it’s Oracle’s entire stack: Solaris x86, Oracle Linux, and Oracle VM. It’s clear Oracle are able to use their database sales .

The second big development has really been overlooked:

World-class Oracle Solaris support on any certified x86 system on the Solaris Hardware Compatibility List (HCL)

That’s right – Oracle are offering Premier Support for Solaris on all certified systems on the Solaris HCL. That’s over 500 server systems – all qualifying for Premier Support.

This is even more relevant when taken in the context of Oracle’s previous comments about increasing investment in SPARC. Why divert development effort to the low x86 end, when you can get your hardware partners to supply the really cheap, low margin gear? Give your customers a neat upgrade path, using the same Solaris OS on a tightly integrated Oracle x86 cluster.

Then supply the customers another upgrade path when they really need to scale – Solaris on SPARC, all the way up to beefy M9000s. The OS is the same, the database is the same, and Oracle can provide customers with a well defined and (relatively) painless upgrade path – which they supply.

This is a real shot in the arm for Solaris x86 and will go a long way to allay the fears of sites who’ve chosen to deploy on x86.

Sun Studio compiler options – a beginners guide Comments Off on Sun Studio compiler options – a beginners guide

Over at Sun’s HPC blog, Thierry Manfe has a nice blog up looking at compiler optimisation flags in the Sun Studio compiler. I try and use Sun Studio when building stuff on Solaris, because not only does it aid performance, but if you know what you’re doing you can use it to really optimise for your processor.

And that’s the problem – there are a whole raft of command line options, and if you’re just starting out you’re presented with a dizzying array of possible optimisations.

Theirry’s post discusses some of the obvious ‘quick win’s you can make, as well as covering their potential downsides. It covers such gems as -fast:

If you are in a rush, you can use the -fast option. What it really does is triggering a set of other options for maximum runtime performance.

Head on over to this page and have a read through the full post – it’s very handy and has some useful tips on building some really optimised code.

New home for orphaned Sun projects Comments Off on New home for orphaned Sun projects

There was always going to be some fallout from the Oracle takeover of Sun – projects that were still in the development phase, technologies that weren’t making enough money – and there have been questions about how the open source casualties would continue.

Izumo Shinohara, a recently ex Sun employee, has setup a new site – Red Giant Phase – which aims to provide a new home for these orphaned projects.

Currently listed are projects like Wonderland, Dark Star, and Project DReaM, amongst others.

It’ll be interesting to see what happens to these open sourced projects now that they’re outside Sun – I wish them all the best.

Sun and Oracle – aftermath of the big event Comments Off on Sun and Oracle – aftermath of the big event

Last week Oracle held a marathon 5 hour webcast session, where they laid out their plans for Sun and their technologies. Sun’s website now redirects to Oracle, and although all the old Sun website links are still live, it’s now Oracle through and through.

The webcast held no surprises, really. As I mentioned previously about the Sun/Oracle merger and Larry’s talk on Oracle’s use of Sun technology, Oracle weren’t going to ditch Sun’s hardware line. The analysts were full of hysterics and gloom, but I’ve yet to meet an analyst who has the slightest clue of what’s going on. They’re paid to make noise and sell ‘research’, not to know what they’re talking about.

As predicted, there’s more investment in Sun’s hardware line, including lots more tasty new CMT processors, and a scaling up of the line to larger multi-socket machines. The high end gear will continue, as will the partnership with Fujitsu. SPARC continues to get a lot of investment and love, and will be a big focus going forwards. Amen to that.

Pretty much all of the software stack will stay and get integrated with Oracle’s offerings. I note with distaste that Oracle’s crappy Internet Directory remains the ‘enterprise’ offering for LDAP and identity management, with Sun’s LDAP products being pushed at smaller deployments. On the OS side, it’s a bit of Linux, and Solaris, Solaris, Solaris – Oracle recognise it’s the best commercial UNIX currently on the market, and that the feature set is unmatched.

Storage lives on, with Sun’s excellent Amber Road Storge 7000 Unified Storage boxes becoming ZFS appliances. Particularly exciting is their integration into OEM – imagine simple management of RMAN backups to ZFS appliances, giving low level snapshots and all sorts of goodness. I can see a lot of places going for that in a big way.

The big question for me was around OpenSolaris. No mention of it at all. It’s Open Source – that particular cat is out of the bag, and it’s not going away. So the question is what sort of effort will Oracle put behind it? Lots of new, OpenSolaris specific features – the new IPS packaging system and the Automated Installer – have potential, but aren’t up to scratch yet, and quite frankly don’t play well with existing Solaris infrastructure.

My bet is we’ll see less effort in re-inventing the wheel, and more focus on making OpenSolaris a more palatable Solaris 11. There’s a big Solaris installed base out there, and the focus on x86 and new features has so far meant that OpenSolaris isn’t really a credible upgrade path.

As I expected when I first heard the news, Oracle are going to be leveraging Sun’s technology and services and own and optimise the entire stack, from the silicon up to the application. This gives them a chance to really tune everything and to go head to head with IBM. May you live in interesting times, as they say.

Obviously there’s more, lots more. Oracle have handily posted up each section of the webcast so you can pick and choose which session you want to watch here. There are also a series of special short webcasts which focus on specific product areas – you can view them all here.

PS: as a side note, Thomas Kurian, who presented the Software Strategy webcast, managed to give one of the dullest presentations I’ve seen. Seriously, that was a really important session, but I almost nodded off a couple of times. Dire.

Top of page / Subscribe to new Entries (RSS)