ActiveMQ 5.10
We are currently looking at the ActiveMQ high performance Journal in
conjunction with the JDBC MySQL adapter with a view to improving the DB
persistence efficiency and performance.
We have it up and running and it has reduced the transactions as expected
however we are looking to get confirmation regarding the Journal log files
and the DB checkpoint functionality.
Therefore could you answer the following questions please:
1.The number of journal log files and size are configured in the
activeMQ.xml file eg the following will create 3 log files 1mb in size
<persistenceFactory> <journalPersistenceAdapterFactory
journalLogFiles="3" journalLogFileSize="1MB"
dataDirectory="${activemq.data}" dataSource="#mysql-ds"
useDedicatedTaskRunner="false"/></persistenceFactory>
2.Every 5 mins the high performance journal will checkpoint to the DB.
3.The 5 mins is not configurable, its part of the code.
4.When the log files are initially created and messages are pushed onto the
broker, messages are written to the journal log files. As mentioned above
every 5 mins the journal checkpoints data to the database. What happen when
the log file gets full, ie in our instance 1MB
5.What happens when the last log file gets full?
6.What happens if a message that is larger than a journal log file is pushed
onto the broker, eg in our instance greater than 1MB?
7.If the AMQ node fails and is then brought back up is the journal flushed,
is this a configurable behaviour?
8.If the AMQ node starts up with a journal that is out of sync with the
underlying DB, what happens, does the journal try to make checkpoints to the
DB that are incorrect etc..?
Please could you provide some insight into the above questions as no real
clear information exists as to how this works..
--
View this message in context:
http://activemq.2283324.n4.nabble.com/ActiveMQ-JDBC-High-performance-Journal-tp4703697.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.
