openqm           début
qu' est ce qu' openqm ?

qu'est ce que les sources ouverts ?

achetez qm aujourd'hui

téléchargements

qu'est qu'il y a de neuf dans les maj récentes ?

support

questions fréquemment posées

l'information de développeurs

applications et outils de développement

politique de confidentialité

liens

contactez-nous
--------------------------

installation
product features
what is the relationship with other mv databases?
product licensing
how do release numbers work?
what is a dynamic file?
keyboard input problems
qmserver problems
the qmclient api
security issues
can i sell my own qm applications?
terminal support
what is a dynamic file?

dynamic files are used to store most data in qm applications. they employ a hashing process to locate records by a mathematical transformation of the record id to a file position. the aim is that a record can be located with just a single disk access regardless of whether the file contains 10 records or 10 million records. in reality, a small proportion of the records in a file may take more than one disk access but the average should be very close to one.

most high performance file systems need periodic retuning to maintain their performance as the data in the files changes. this usually requires users to be logged off the system for some time. dynamic files are automatically restructured in very small steps during normal use, effectively spreading the cost of retuning very thinly over the normal operation of the file.

although there are configuration parameters that can be used to adjust the way in which dynamic files work, most users simply leave them at their defaults and let the system sort itself out.

for more information about the inner workings of dynamic files, visit the technical notes page for a detailed description.