activemq-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Fuzzo <>
Subject Re: Derby journaledJDBC persistence adapter never checkpoint!
Date Wed, 24 Sep 2008 12:56:50 GMT

I've tried with but without

With help from Derby forum, the cause may be a long-running transaction that
is never committed and it blocks derby's checkpoint.
I thought it may be ActiveMQ database lock strategy but i have not any probe

In fact, if i close all ActiveMQ instances the Derby db isn't closed
correctly with a checkpoint. 

If i reopen the same DB with an SQL client (like ij, SQirreL, ...) the
recovery mode is called and a checkpoint is done immediatly, deleting all
log files except the last one!!!

See also
this thread  for details...

I hope someone can help me with this issue... Thanks!

Fuzzo wrote:
> Hi all!
> I'm using ActiveMQ with journaledJDBC persistent adapter, configured like
> this:
> <persistenceAdapter>
>     <journaledJDBC journalLogFiles="5"
> dataDirectory="\\path-to-shared-folder\activemq\db\" />
> </persistenceAdapter>
> At this location there are:
> derbydb <dir>
> journal   <dir>
> derby.log
> Inside derbydb/log there are a lot of files in format:
> logn.dat (~1MB) where n is between 1 and, after a day, 892 for a total of
> 904MB of files!
> Are these files the log of derby? Why it does not do the default
> checkpoint after 10MB (ten log files)? There is a way to configure this?
> The disk space on shared NAS isn't unlimited and ActiveMQ in my project is
> intended for running 24/7!
> ActiveMQ runs with 5 topics and 10 subscriptions, if this can help you to
> solve my problem!
> Thanks a lot!
> P.S.: In the Java Service Wrapper configuration file used for AMQ, there
> is this parameter Can it
> affect derby checkpoint?

View this message in context:
Sent from the ActiveMQ - User mailing list archive at

View raw message