Traffic Data Archival and Cleanup
By default, raw traffic will be accessible at the scheduled
interval (for example, 5 minutes) for the last 35 days. Beyond the
35 days, there will be only one data point per day. Accumulated traffic
data is stored and gzipped under /u/wandl/data/traffic.archivefor a certain number of days (35 by default).
If this setting uses up too much disk space, the parameter archiveCapacity
(35 by default) can be edited in /u/wandl/db/config/agg.xml to a smaller
number. For proper display of historical traffic charts from the web,
any changes to this value in the agg.xml file should also be made
to the parameter archiveCapacity in /u/wandl/web/wandl/WEB-INF/web.xml.
If disk space is running low, a cron task can be setup to remove or backup old archive files. To set up a cron job, switch to root user. Run the command “crontab -e”. An example entry for removing old archive files at 1:00 AM every day (before aggregation, which is usually set to run at 2am) is as follows:
0 1 * * * rm /u/wandl/data/traffic.archive/*Archive files will contain the raw data at the scheduled interval (for example, 5 minute default). After being archived, this data can be accessed via mysql commands.
By default traffic data older than 100 days is purged.
Traffic aggregation logs are stored in /u/wandl/log:
- agg.log.n:Logging output from the Aggregator. Edit log levels in /u/wandl/db/config/agg.log.properties file. This rotating log is 500KB max.
- agg.msg:Includes stdout and stderr output. All exceptions will appear here.
- aggTASK.log.n:Data flush capture
- aggDB.log.n:Database access log with full trace of all SQL queries
To change log levels, edit /u/wandl/db/config/agg.xml
