Buildfarm build_test script on an embedded device.

[Andrew Bartlett]

On Tue, 2011-10-04 at 22:57 -0500, Christopher R. Hertel wrote:

> :
> > If you can find a box that is around 10x more powerful (in ram and CPU)
> > that might help.  I'll know more when I see the logs.
> 
> My gut reaction is that it's primarily the RAM and I/O speed.  The Dockstar
> has only 128MB of RAM plus a generous amount of swap, but the swap is on the
> same drive as the build environment and it's just a USB-connected 2.5"
> laptop drive.  Not much in the way of I/O speed.
> 
> Logs on the way...
> (Separately)

The logs indicate that the timelimit fired.  This indicates to me that
it is the build that takes the bulk of the 24 or 14 hours (I guess the
difference might be due to ccache).

There are also other timeouts, which make this much less useful for
actually performing portability work with. 

As I said, let's first ensure we are clean on Arch x64 or x86, then
let's see what we can do to slim stuff down - the same efforts that are
already planned for better scalability will also help at the low end,
and real world use cases won't torture the device in the way our
testsuite does. 

Andrew Bartlett

-- 
Andrew Bartlett                                http://samba.org/~abartlet/
Authentication Developer, Samba Team           http://samba.org