(no subject)

[Bob Sneidar]

Hi Ruslan.

2 things weigh against your assessment. First, one of the databases is  
remote and it happens to be the one I am querying 1 record at a time.  
The second thing is that to my knowledge you cannot create joins  
between 2 tables in different databases. The source database is a  
local file based SQLite database to further complicate matters.

I want to employ a method that could work with databases of 1000  
records or 1,000,000 records. To do that means I cannot risk memory  
overflow in runrev by huge query results. For this particular  
application I would never hit that limit, but what if I want to take  
the functionality of my data import module and make it available by  
itself?it needs to work in all situations or not at all. That's how I  
think.

My goal here is to import/merge data from a Spiceworks database of  
devices, software and other things. But in the future it could be  
anything. I am learning the limitations of querying remote SQL  
servers. For this application I think a local SQL server would be much  
better, but for the dev phase where I am out and about with my laptop  
on-rev will suffice.

Bob Sneidar
IT Manager
Calvary Chapel CM
Sent from iPhone

On May 21, 2010, at 22:58, Ruslan Zasukhin <ruslan_zasukhin at valentina-db.com 
 > wrote:

> Еexcuse me but this still sounds like a very bad result.
> SQL dbs can make joins of thousands records in say .001 sec