How to Configure Statistics Processing and Maintenance

Statistics data processing and maintenance is configured on the Barracuda Firewall Control Center on a global, range, or cluster specific basis. In the Cook Settings table, add new, or edit existing settings to change how the dstats processes and maintains statistics for specific data types. Modify the default settings to suit your specific statistics requirements.

Statistic cook settings do not apply to how statistics files from the Barracuda Firewall Control Center system itself are processed.

Configure Global Statistics Settings

Go to CONFIGURATION > Configuration Tree > Multi-Range > Global Settings > Statistics Cook Settings.
Click Lock.
From the Corrupted Data Action list, select one of the following options to specify how dstats handles a corrupted DB file:
- Delete – Deletes the corresponding DB file (default). When statistic files are deleted, they are only deleted from the Control Center; the files are not deleted from the CloudGen Firewalls.
- Archive – Moves the DB file to a lost+found directory.
When a corrupted data file is recognized, the Corrupted Data File [150] event is triggered.
Edit the global statistics cook settings. For more information see the Cook Settings section below.
Click OK.
Click Send Changes and Activate.

Configure Range-Specific Statistics Settings

Go to CONFIGURATION > Configuration Tree > Multi-Range > your range > Range Properties.
Click Lock.
To enable specific cook settings for all clusters in the range, enable Own Cook Settings.
Click Send Changes and Activate.
To edit the cook settings for the range, open the Statistics Cook Settings page (CONFIGURATION > Configuration Tree > Multi-Range > your range > Range Settings).
Configure the settings as described in Configure Global Statistics Settings.
Click Send Changes and Activate.

Configure Cluster-Specific Settings

Go to CONFIGURATION > Configuration Tree > Multi-Range > your range > your cluster > Cluster Properties.
Click Lock.
To enable specific cook settings for all systems in the cluster, enable Own Cook Settings.
Click Send Changes and Activate.
To edit the cook settings for the cluster, open the Statistics Cook Settings page (CONFIGURATION > Configuration Tree > Multi-Range > your range > your cluster > Cluster Settings).
Configure the settings as described in Configure Global Statistics Settings.
Click Send Changes and Activate.

Cook Settings

Setting	Descripton

Setting	Descripton
Settings for	The software module that generates statistics data. Optionally, you can select Pattern Match to define a file pattern that should apply to the statistics data settings. Generally, you do not need to use directory patterns because the default statistics settings meet most statistics requirements. If you do use directory patterns, make sure that they do not interfere with the module settings. For a specific data type, use EITHER the module OR directory pattern settings. dstats works through the configured instances successively and omits directory patterns that apply to directories that have already been processed. Additionally, for clearly arranged management, place directory patterns at the end of the configuration file.
Directory Pattern	Pattern-Match settings apply to statistics files that are available in the sub-folders of `/var/phion/stat` . Enter full folder names or wildcards (? and ). The question mark wildcard (?) indicates a single character, and the asterisk wildcard () indicates an arbitrary number of characters. Example Pattern: To include all firewall statistics files that start with 'conn' from all servers that start with 'S' in ranges 1 and 2, enter the following pattern structure: Actual file structure: `/var/phion/stat/1/S1/service/FW/conn<xxx>` `/var/phion/stat/1/S2/service/FW/conn<xxx>` `/var/phion/stat/2/S3/service/FW/conn<xxx>` Directory pattern: `/S?/service/FW/conn` Avoid using too openly defined patterns that span multiple folders, such as /service//*. If you do use patterns that span multiple folders, be aware of their effect and put them at the bottom of the list.
Type: Time	Options in this section only apply to Time statistics. For example byte (Time for Dst) and conn (Time for Src). Resolution 1h after (days) – Number of days after which the granularity of time statistics is increased to one hour. Data that is more recent than the specified number of days will not be affected. Resolution 1d after (days) – Number of days after which the granularity of time statistics is increased to one day. The period between cooking from hour to day granularity must be at least 2 days. If only 1 day is specified, it will result in a summary offset for hourly granularity of 0 days per instance. This will lead to an error message in the dstat log file that is similar to the following: Cannot create, file byte.hour_tot<cookInstStartTS> exists already. Delete Data after (days) – Number of days after which time statistics are deleted.
Type: Top	Options in this section only apply to Top statistics. For example, byte (Top Dst) and conn (Top Src). Condense after (days) – Number of days after which top statistics are merged into larger temporal bins. Data that is more recent than the specified number of days will not be affected. Delete Data after (days) – Number of days after which top statistics are deleted. Resolution – Specifies if top statistics are condensed to a week or month. Only change this setting while the system is unproductive. Otherwise, statistics will not be displayed properly because of incomplete cook instances.