Hi Paul,

What ingres version and patch?

Actually the printqry looks pretty good. What do you think is the
problem there?

How many rows are in this table?

The copy has mentioned minpages=10, but hasn't included a
row_estimate. So there shouldn't be many rows, but if there are then
you may be seeing the program grind to a halt due to overflow in the
hash.

Martin Bowes

--
Random Duckman Quote #97:
Duckman: You got anymore of those glues?
Cornfed: Clues.
Duckman: I'm one letter off, crucify me!

Hi Martin - thanks for the reply.

Ingres Version = 2.0/9808
Rowcount = 3307

The problem with the printquery is that the last SQL query (the copy
statement) has a Query Text and a Query Send Time, and then instead of
a Query Response Time, it has another Query Text and Query Send Time,
both the same as before. This happens again and again whenever I rerun
the script, and its not something I've seen before.

I'll change the minpages value if needs be - I'm just going with what
copydb created for me based on the data in the table.

Paul

Self wrote:
> Hi Paul,
>
> What ingres version and patch?
>
> Actually the printqry looks pretty good. What do you think is the
> problem there?
>
> How many rows are in this table?
>
> The copy has mentioned minpages=10, but hasn't included a
> row_estimate. So there shouldn't be many rows, but if there are then
> you may be seeing the program grind to a halt due to overflow in the
> hash.
>
> Martin Bowes
>
> --
> Random Duckman Quote #97:
> Duckman: You got anymore of those glues?
> Cornfed: Clues.
> Duckman: I'm one letter off, crucify me!

Paul,

Is there anything written to the errlog.log when you run this copy
statement?

John

Hi Paul,

> Hi Martin - thanks for the reply.
>
> Ingres Version = 2.0/9808
> Rowcount = 3307

That row_count is pretty small! No matter what this shouldn't hang.

Try:
Add a row_estimate=3307 to the copy with clause.
* Alter the script to do the copy before the modify.
* Save the ascii data file aside for the moment, and then try this as a
binary copy.
* Check (using a text editor) that the data in the ascii data file is
sensible

Marty

>
> The problem with the printquery is that the last SQL query (the copy
> statement) has a Query Text and a Query Send Time, and then instead of
> a Query Response Time, it has another Query Text and Query Send Time,
> both the same as before. This happens again and again whenever I rerun
> the script, and its not something I've seen before.
>
> I'll change the minpages value if needs be - I'm just going with what
> copydb created for me based on the data in the table.
>
> Paul
>
> Self wrote:
> > Hi Paul,
> >
> > What ingres version and patch?
> >
> > Actually the printqry looks pretty good. What do you think is the
> > problem there?
> >
> > How many rows are in this table?
> >
> > The copy has mentioned minpages=10, but hasn't included a
> > row_estimate. So there shouldn't be many rows, but if there are then
> > you may be seeing the program grind to a halt due to overflow in the
> > hash.
> >
> > Martin Bowes
> >
> > --
> > Random Duckman Quote #97:
> > Duckman: You got anymore of those glues?
> > Cornfed: Clues.
> > Duckman: I'm one letter off, crucify me!
>
> _______________________________________________
> Info-ingres mailing list
> Info-ingres@cariboulake.com
> http://mailman.cariboulake.com/mailm...py/info-ingres
>

Haha - problem solved!

>Is there anything written to the errlog.log when you run this copy
>statement?

Yes, there was:

>Wed Oct 26 09:30:43 2005 E_DM9711_SR_OPEN_ERROR Error occurred opening
>sort work file -- path = /NIHE/tmp/ingres/work/default/pregion3, file = g4aaabtk.o00.

Basically, the Ingres temporary work directory for the database in
question didn't exist. We lost this area of the hard disk some time
ago, and have been creating the directories for the different databases
one by one as we need them. We don't use this particular database very
often, but I hadn't realised that there might have been a problem here
(particularly since I had done another table change on that database
yesterday, without errors).

I've created the work directory, and the copy.in script works fine now
- thanks for the help, guys.

:-)

Paul