MCslp Coalface

Comparing 32-bit/64-bit MySQL on OpenSolaris

I’ve been working with the folks working on OpenSolaris for a few months now providing advice and input on getting MySQL and the connectors (C/ODBC and C/J) installed as a standard component. Having got the basics in, the team are now looking at adding both 32-bit and 64-bit packages.

The question raised at the end of last week was whether OpenSolaris should enable 64-bit builds by default in 64-bit installations, and whether there was a noticeable performance difference that would make this worthwhile.

I did some initial tests on Friday which showed that there was a small increase (10-15%) of the packaged 64-bit installations over 32-bit under x86 using snv_81. Tests were executed using the included sql-bench tool, and this was a single execution run of each package for 5.0.56. Transactions are missing because I hadn’t enabled transactions in the tests.

There are some significant differences there (like the 55% increase on SELECT speeds, for example), but a single execution is never a good test. Also, it’s unclear whether the differences are between the compilations, the platform or just pure coincidence. This requires further investigation.

As a coincidence, Krish Shankar posted these notes on using SunStudio 11 and SunStudio 12 and the right compiler flags to get the best optimization.

I decided to do 10-pass iterations of sql-bench and compare both 32-bit and 64-bit standard builds, the 32-bit standard builds against Krish’s optimizations, and finally 32-bit and 64-bit optimized builds.

Some notes on all the tests:

• All builds are 5.0.56
• All tests are run on SunOS 5.11, snv_81
• Tests are executed on the same OS and machine running in 64-bit. The SPARC tests are on an UltraSPARC IIIi@1.28GHz Workstation with 1GB RAM; x86 are on a Dell T105, Opteron 1212 with 4GB RAM. Of course we’re not comparing machine speed, just 32-bit binaries over 64-bit.
• All results are in seconds; lower values mean faster performance.
• In all tests I’m using the built-in defaults (i.e. no my.cnf anywhere) so as to simulate a standardized installation.

Let’s first look at x86 and the 32-bit standard and 32-bit optimized builds:

The standard build uses gcc instead of SunStudio, but I don’t get the same performance increases that Krish saw - in fact, I see reductions in performance, not improvements at all. I’m going to rebuild and retest, because I’m convinced there’s a problem here with the builds that I’m not otherwise seeing. I certainly don’t expect to get results that show a 27% reduction in insert speed. That said, a 10% big-table increase is interesting. I’ll redo these builds and find out if the slow down is as marked as it here.

Here’s the comparison for standard builds between 32-bit and 64-bit standard builds on x86:

There are some incredible differences here - more than 50% increase in SELECT, and 30% for the big-tables test show that there is some advantage to having the 64-bit builds on x86 enabled.

Unfortunately I’ve had problems with the 64-bit optimized builds on my machine, so I haven’t completed optimized test comparisons.

On SPARC, Sun Studio is used as the default compiler, and the standard 32-bit and 64-bit show little difference:

Overall, a pretty insignificant difference here.

Now let’s compare the standard and optimized builds using Krish’s flags on SPARC:

The tests here show little significant difference between the standard and the optimized builds, although 6-7% would probably be enough to prefer an optimized build if you wanted to build your own.

Now let’s compare the optimized, Sun Studio 12 builds running in 32-bit and 64-bit:

The differences are virtually non-existent, and taking the reductions and increases in performance overall, there’s probably little difference.

The overall impression is that on x86 the improvement of 64-bit over 32-bit is significant enough that it’s probably a good idea to make 64-bit the default. On SPARC, the difference in the optimized builds is so slight that for compatibility reasons alone, 32-bit would probably make a better default.

I’ll probably be re-running these tests over the next week or so (particularly the x86 so I can get a true comparison of the 64-bit optimized improvements), and I’ll try the T1000 which I upgraded to snv_81 over the weekend, but I think indications are good enough to make a reasonable recommendation of 64-bit over 32-bit.

Can a website have a box?

I’ve been sent a couple of books to review for O’Reilly, which I’ll get to soon, but I’m intrigued by the text on the front of the book.

Both are for websites - Wikipedia and Facebook - and are both part of the Missing Manual series.

I’m confused though - both also quote the Missing Manual mantra:

The book that should have been in the box

Can a website have a box?

Acorn, Pixelmator and Iris alternatives to Photoshop

I’ve had Acorn in my list of applications to review for months, and I’ve only just got round to it.

I wish I’d got there earlier. Acorn is quick and powerful, and that’s because it employs your GPU to do soe of the processing, and it includes a number of filters (based on OS X’s CoreImage interface), all of which is wrapped up into a nice little application. If you can’t find what you want, there are ObjectiveC and Pythong plugin interfaces, but I haven’t investigated it yet.

Of the alternatives, the most talked about is Pixelmator, closely followed by Iris.

Pixelmator is a closer approximation to the way that Photoshop operates, and in some respects I prefer the functionality and the feel of Pixelmator if I was looking for a Photoshop replacement, but there are other elements I don’t appreciate. The flashy graphics and animations when you do different operations seem superfluous to me.

There are nice touches in both applications - the stamp tool in Pixelmator is particularly good (although I prefer Photoshop), while in Acorn the crop and select tools provide much better feedback during the select operation than even Photoshop.

Iris is less polished, but shows some promise. There are some annoying oddities (I used 1.0b2, 367), like the image opening at pixel resolution, rather than being scaled to screen size, and the lack of specialized cursors can make identifying what you are doing and the potential effects of that process difficult, but the image editing and manipulation is very quick (particularly on stamp and touch up operations). It is, however, a bit memory hungry at the moment.

Any of these solutions would make a good alternative Photoshop and Photoshop Elements if you don’t want to go down the Adobe route.

Of these I currently prefer Acorn - it’s small and lightweight and the interface feels much more polished and easy to use. Certainly I’d consider it as an alternative to the larger packages on a laptop if you wanted something while you were traveling. I can’t get by without Photoshop because of the image scanning and editing I do, but occasionally I want something more extensive than Preview when I’m on the move.

Of course, this could change - all of these tools are being actively developed and so it’s likely that there will be some leapfrogging along the way.

Migrating from WSRF to WSRT

If you’ve been reading the main MCslp blog you will know that I’ve been writing a series on writing a grid solution using WS-RT.

WS-Resource Transfer is an improved version of the WS-Resource Framework (and the associated WS-ResourceLifetime and WS-Transfer). Basically, the developers of WS-RT looked at WS-Transfer and WSRF, and added the functionality found in WSRF missing in WS-Transfer to WS-RT. WS-RT, for me, is basically what WS-RF should have been all along.

If you know WS-RF and want to start using WS-RT, check out Migrating from WSRF to WSRT.

For my WS-RT series, check out the MCslp Grids category.

An introduction to Eclipse for Visual Studio users

I’m seeing more and more people moving to Eclipse as a development platform, even those Windows users who have traditionally used Visual Studio. As an Eclipse user for quite a while now I’m often asked how good it is, or how to use it.

Of course, telling people to simply try it out isn’t enough. Many people just don’t get Eclipse and cannot understand or translate the skills and experience they already have to the Eclipse environment. That’s where An introduction to Eclipse for Visual Studio users can help.

It’s a quick overview of the fundamentals of Eclipse from the perspective of a Visual Studio user. For a more in depth examination, there’s a tutorial Eclipse for Visual Studio developers, and another on migrating your applications from VS to Eclipse: Migrate Visual Studio C and C++ projects to Eclipse CDT.

I can recommend any (or indeed all) of these.

Getting Best out of MySQL on Solaris

I’m still working up some good tips and advice on MySQL on Solaris (particularly the T1000, the new x86 based servers like the X4150 and ZFS, amongst other things), but until then I found Getting Best out of MySQL on Solaris while doing some research.

With the latest OpenSolaris builds (b79, from memory) we now have MySQL built-in, and I worked with the folks on the OpenSolaris Database team to get some reasonable configurations and defaults into the system. MySQL 5.1 and 64-bit support is currently going through the process and will be a in future build.

I’ve also been working with the DTrace people to improve the DTrace support we have in MySQL (documentation will go live this week, I hope). MySQL 6.0.4 will have some basic DTrace probes built-in, but I’ve proposed a patch to extend and improve on that significantly.

We’re in the process of updating the documentation and advice on Solaris (and OpenSolaris) installations and layout too, which is itself part of a much larger overhaul of the installation and setup instructions for all platforms.

Intute Science, Engineering and Technology - Full record details for Managing a grid, part 1 : network and infrastructure

One of the nice things that happens when I’m writing my articles is that while researching the content, I come across a blog post or other site that has picked up one of my articles and added it to some catalogue or other.

Here’s an example I just found: Intute Science, Engineering and Technology - Full record details for Managing a grid, part 1 : network and infrastructure.

A more detailed look shows it’s not the only article they have of mine in their catalog (see this search). There are lots of useful articles here if you are still interested in grid tech and want some more background information.