Many many thanks for your help! I've tried to do this: - shut down ActiveMQ (derby checkpoint NOT done) - opened the db with ij (slow, for recovery mode i suppose, checkpoint DONE)
The final size is now < 5 MB! :) Now i'm trying to open derby debug query log file (1.17GB...) looking for queries that may block checkpoint... See you soon! :) Knut Anders Hatlen wrote: > > Fuzzo <[EMAIL PROTECTED]> writes: > >> Thanks for the answer. >> >> I've read these instructions in documentation but the Derby instance i >> use >> is embedded in ActiveMQ, not in my own application, and i don't wont >> modify >> the source code! :) > > You don't need to modify the source code to try it. Just start the ij > command line tool, connect to the database and invoke the > statements. That would give a better indication of what the problem is. > >> My questions are: >> - can i override any ActiveMQ Derby properties to force default >> checkpoint >> (after 10MB of logs)? > > The default settings are: > > derby.storage.logSwitchInterval=1048576 > derby.storage.checkpointInterval=10485760 > > That is, switch to another log file when the current log file exceeds 1 > MB, and checkpoint when you have 10 MB of log. > >> - why ActiveMQ shoud enable log archiving mode? > > No idea, it's just a guess. Checking if disabling log archiving helps, > would be a quick way to find out if this is the case. > > Another possibility, also mentioned by others, is that there is a > long-running transaction or a transaction that the application has > forgotten to commit or abort. The checkpoint will keep all the log back > to the start of the oldest active transaction, so that could explain the > size of the log. In that case, most of the log should be deleted when > the database is rebooted. Does that happen? > > -- > Knut Anders > > -- View this message in context: http://www.nabble.com/Embedded-Derby-and-.dat-files-tp19574483p19644779.html Sent from the Apache Derby Users mailing list archive at Nabble.com.
