On 30/04/14 11:43, integrity wrote:

Hi,

the elapsed time is 6:30.39, means 6 minutes 30.39 seconds.

No, you have to look at the inputs/output and pagefaults numbers
in both runs. This will show if you run hot/cold

Having concurrent load on your machine also may have an impact.
For example, browsers can take a substantial amount of memory
and cpu away from database processing.

To understand this, you can look at the output of 'uptime' and 'top'

Regards, Meng

------------------ Original ------------------

From: "Martin Kersten";<Martin.Kersten@cwi.nl>;

Date: Wed, Apr 30, 2014 05:38 PM

To: "users-list"<users-list@monetdb.org>;

Subject: Re: 回复： why the same procedure (in shell or in sql) cost so differenttime?

On 30/04/14 11:27, integrity wrote:
> Hi,
...
>
> On 30/04/14 10:59, integrity wrote:
> > Hi All,
> > I want to execute a batch of small sql stored procedures, for
> convenience, i put them totally in another big stored procedure:
> gwac_uniquecatalog, it cost 58 minutes:
> > #!/bin/bash
> > time mclient -d mydb -s "CALL gwac_uniquecatalog(99);"
> > real 58m3.518s

> >
> > The time is so long that I tried to execute the small stored
> procedures one by one in SHELL, it only cost 6.5 minutes
> > /usr/bin/time ./gwac_uniquecatalog.sh 99
> > 2.82user 3.25system 6:30.39elapsed 1%CPU (0avgtext+0avgdata
> 10880maxresident)k
> > 0inputs+0outputs (0major+980011minor)pagefaults 0swaps
> >
What were the numbers of /usr/bin/time for the long run?
_______________________________________________
users-list mailing list
users-list@monetdb.org
https://www.monetdb.org/mailman/listinfo/users-list
.
_______________________________________________
users-list mailing list
users-list@monetdb.org
https://www.monetdb.org/mailman/listinfo/users-list