Slow Query only with some values??????

Question

I have this tables:
Vehicles: id_vehicle, id_customer, registration_number, model, ... (2.000.000 of records) (1.500.000 distinct id_customer)
Insurance-Customers-Vehicles: id_insurance, id_customer, id_vehicle ... (4.500.000 of records)
Insurance: id_insurance, date_insurance, ...
Customers: id_customer, name, surname, ..
QUERY:
select
vh.id_vehicle,
vh.id_customer,
vd.model,
vd.registration_number
from Insurance-Customers-Vehicles vh
left join Vehicles v on (vh.id_vehicle = v.id_vehicle and vh.id_customer = v.id_customer)
inner join Vehicles vd on (vh.cod_vehiculo = vd.cod_vehiculo)
where vh.id_customer = 123456
and vh.id_vehicle is null
I execute this query because I want to obtain Vehicles that have been associated with a customer but I don’t want to obtain the vehicle-customer on Insurance-Customers-Vehicles table if it’s now associated to the vehicle-customer on vehicles table.
This query is very fast at 99% of the time (less than 1 second), but when a client has many vehicles associated (more than 50) on Insurance-Customers-Vehicles table the query is very slow and takes between 10 and 20 seconds ...
Insurance-Customers-Vehicles
Please, How can I improve it?????
Thanks in advance!!!!!
Operation     Object Name                              Rows     Bytes     Cost     Object Node     In/Out     PStart     PStop
SELECT STATEMENT Optimizer Mode=CHOOSE                         3             12                                          
  NESTED LOOPS                                        3       498       12                                          
    FILTER                                                                
      NESTED LOOPS OUTER                                                                
        TABLE ACCESS BY INDEX ROWID     Insurance-Customers-Vehicles     3       408       6                                          
          INDEX RANGE SCAN     IND_INSCUVEH_ID_CUSTOMER          3             3                                          
        TABLE ACCESS BY INDEX ROWID     VEHICLES               1       13       2                                          
          INDEX UNIQUE SCAN     PK_VEHICLES                    1             1                                          
    TABLE ACCESS BY INDEX ROWID     VEHICLES                    1       17                                                
      INDEX UNIQUE SCAN     PK_VEHICLES                         1             1                  

Answer 1

I think there is something logically wrong with this query.
You left join the vehicle table, however you already know that id_vehicle is NULL.
I'm not sure what it should achieve, maybe you should replace this join with an EXISTS sub select, maybe you can remove this join completely.
Try to find an example where your query gives a different result than this one. And explain me why the result is different.
select vh.id_vehicle,
      vh.id_customer,
      vd.model,
      vd.registration_number
from Insurance-Customers-Vehicles vh
inner join Vehicles vd on (vh.cod_vehiculo = vd.cod_vehiculo)
where vh.id_customer = 123456
and vh.id_vehicle is null;Message was edited by:
Sven W.