Re: [HACKERS] Parallel regress tests (was Re: FOREIGN KEY andshift/reduce)

Tom Lane wrote:

> Thomas Lockhart <lockhart(at)alumni(dot)caltech(dot)edu> writes:
> >> Kind of like a simple semaphore. Isn't that what you are getting at?
>
> > LISTEN/NOTIFY might be useful for this...
>
> LISTEN/NOTIFY would be one of the things we were trying to *test*.
> It's not helpful for checking intra-transaction-block behavior anyway,
> since notifys are only sent at commit and only heard between
> transactions.
>
> I like the tcl/expect idea a lot. We'd have to upgrade our tcl
> interface to support asynchronous queries (ie, send query then do other
> stuff until answer arrives); AFAIR it can't handle that now. But that'd
> be well worth doing anyway...

That would be really nice, especially because Tcl is my
preferred tool language. But I wouldn't use libpgtcl, instead
it should control a number of psql sessions over command
pipelines. These can already handle asynchronous actions
across all supported platforms.

I'll build the first version of the tool using standard Tcl,
and just keep in mind that the command language must be
easily implementable with lex/yacc so someone can convert it
later into a C version.

Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#========================================= wieck(at)debis(dot)com (Jan Wieck) #