[Bioclusters] Any issues porting applications to OS X?

[Chris Dagdigian]

For apps where source code is available OS X has been generally very 
easy to deal with.

There are, however, commercial apps that are released binary only for 
Linux (and usually only certain flavors of Linux). Two companies who 
sell Linux products into the informatics, chemistry or molecular 
modeling spaces are Schrodinger and Accelrys but I'm sure people can 
come up with more concrete example.

My personal feeling right now (probably at odds with my partners maybe 
:) is that Linux on X86 or X86_64 is a better generic or 
widest-possible-utility computing platform.

The decision to go with an alternative (UltraSparc, G5, whatever...) 
should be based on actual benchmarks or local documented need.

Apple OS X clusters make sense for two distinct groups IMHO:

1. People who have real benchmarks that show G5/darwin as being the 
best/fastest platform for the apps that they most care about. There are 
a number of these as the G5 platform is really really slick.

2. People who like OS X, have lots of inhouse experience and prefer for 
operational overhead to manage their cluster(s) using the same 
technology and tools they already use inhouse.

Yellowdog Linux on G5 is a strange one. Probably best for people who 
have a known set of apps that run really fast on G5 and don't want or 
like Panther. I would not recommend this approach for people wanting a 
general purpose platform -- stick with Panther on the G5 unless you have 
a very specific reason not to.

Linux vs Apple cluster managability is a red herring -- both are easy to 
do. Apple's OS X server OS ships with a number of tools that make 
managment pretty easy. Linux has the same capabilities through the base 
OS itself or via open source tools.

We generally have various bits of Apple kit floating through our 
office/lab at any given time. The G5 Xserves have been shipped back 
sadly but we still have dual-G5 towers running Panther.

On a case by case basis we may be able give SSH login access to a G5 
system to people who are investigating porting or benchmarking issues. 
Feel free to drop me a line privately if this would be of assistance.


-Chris
bioteam




Christopher Porter wrote:

> 
> We're in the market for a cluster; most of our options are Xeon/Linux, 
> but one is a cluster of XServe G5s running OS X. We're going to run some 
> benchmarks to see how the performance compares, but some in of our group 
> have expressed concern that 'the vast majority bioinformatics software 
> is developed on Linux', and 'there may be a long time lag before new 
> software is available on OS X'.
> 
> I have never had problems getting software I need to run on OS X, but I 
> wondered if anyone can provide me with examples of applications that 
> won't run on OS X, or are Linux only (only binaries released & no source 
> available).
> 
> This is only one of the criteria we're judging on, and the performance 
> comparison will be interesting. Any insights on this issue would be 
> extremely useful, though.