Hello to all



I found some new info that might help you (in order to help me).



I restarted the server and I loaded only the application where uses the firebird. No other terminal was connected to server (I am pretty sure because the company was closed). And… tada… the firebird reaches the 50% of cpu immediately and the server was very slow.



The version of the application where the customer has, has a minor bug: importing records from ascii files, updates the records even they have the same data. This bug is solved on next versions because we found on another customer that the file size of the database was became too large with no logical reason.



So I am wondering… if this bug that exists now on this customer, triggers firebird to do some kind of maintenance the database file and for this reason firebirds uses the 50% cpu (dual core machine).



Here, is the gstat of the database at it is running now, after the cpu peak, (with no backup restore maintenance):



Database header page information:

Flags 0

Checksum 12345

Generation 1977919

Page size 16384

ODS version 11.0

Oldest transaction 1977860

Oldest active 1977861

Oldest snapshot 1977856

Next transaction 1977902

Bumped transaction 1

Sequence number 0

Next attachment ID 0

Implementation ID 16

Shadow count 0

Page buffers 0

Next header page 0

Database dialect 3

Creation date Sep 22, 2009 9:55:03

Attributes force write



Variable header data:

Sweep interval: 20000

*END*





dennis



[Non-text portions of this message have been removed]