Slow regression tests on windows

Gevik Babakhani wrote:

I am trying to run regression tests on both windows and RH.
It looks like that the tests on windows run slower than linux
using two machines with same hardware config.

Is this known?

We need far more information than this before we can say much sensibly,
I think.. For example, what configure flags you are using, what
compilers, etc.

From what I can see MinGW regression is somewhat slower than MSVC (and
even than Cygwin) on my buildfarm VM that runs all three, even though
the latter two are rate limited by MAX_CONNECTIONS.

The regression tests really aren't performance tests, though.

cheers

andrew

On Mon, Nov 05, 2007 at 08:43:01AM -0500, Andrew Dunstan wrote:

Gevik Babakhani wrote:

I am trying to run regression tests on both windows and RH.
It looks like that the tests on windows run slower than linux
using two machines with same hardware config.

Is this known?

We need far more information than this before we can say much sensibly,
I think.. For example, what configure flags you are using, what
compilers, etc.

In general, the regression tests are slower on windows, yes. It's because
there's a lot of processes spawned on both client and sderver side..

(Heck, in generall, *everything* in pg is slower on Windows than on Linux
with the same hardware config)

From what I can see MinGW regression is somewhat slower than MSVC (and
even than Cygwin) on my buildfarm VM that runs all three, even though
the latter two are rate limited by MAX_CONNECTIONS.

Uh, you're saying MSVC and Cygwin somehow differ from MingW?

//Magnus

Magnus Hagander wrote:

From what I can see MinGW regression is somewhat slower than MSVC (and
even than Cygwin) on my buildfarm VM that runs all three, even though
the latter two are rate limited by MAX_CONNECTIONS.

Uh, you're saying MSVC and Cygwin somehow differ from MingW?

I'm saying my buildfarm members differ from each other.

In fact, on the last run of each, MinGW was by far the slowest and
Cygwin the fastest in the "make check" step. But there could be any
number of reasons for that, including extraneous activity on the VM host
that could have slowed the whole VM down.

Anyway, comparing regression test speeds is probably not very productive.

cheers

andredw

On Mon, Nov 05, 2007 at 11:01:26AM -0500, Andrew Dunstan wrote:

Magnus Hagander wrote:

From what I can see MinGW regression is somewhat slower than MSVC (and
even than Cygwin) on my buildfarm VM that runs all three, even though
the latter two are rate limited by MAX_CONNECTIONS.

Uh, you're saying MSVC and Cygwin somehow differ from MingW?

I'm saying my buildfarm members differ from each other.

Ah. I thought you meant we had some builtin cap in the regression tests,
which is what confused me.

In fact, on the last run of each, MinGW was by far the slowest and
Cygwin the fastest in the "make check" step. But there could be any
number of reasons for that, including extraneous activity on the VM host
that could have slowed the whole VM down.

Yeah. It does surprise me a lot that cygwin should be fastest, really.

Anyway, comparing regression test speeds is probably not very productive.

Agreed.

//Magnus

Magnus Hagander wrote:

In fact, on the last run of each, MinGW was by far the slowest and
Cygwin the fastest in the "make check" step. But there could be any
number of reasons for that, including extraneous activity on the VM host
that could have slowed the whole VM down.

Yeah. It does surprise me a lot that cygwin should be fastest, really.

Don't read anything into it. The VMware clock is quite unreliable, in my
experience.

cheers

andrew