- Home
- Unravel 4.5 Documentation
- Advanced Topics
- Unravel properties
- Disk monitoring properties
Disk monitoring properties
File system related rules (disk usage) are evaluated separately.
Property/Description | Set By User | Unit | Default |
|---|---|---|---|
com.unraveldata.monitoring.fs_rules.eval.interval Frequency to evaluate file system rule. Sets the timing interval for partition (volume) and folder (directory) monitoring. 0 disables the rule. | s | 60 |
A list of partitions to be monitored. Once partition usage exceeds given threshold (configured as high watermark limit) an email alert is sent. Another email is only sent when the disk usage goes below low watermark limit and above high watermark limit again.
Property/Description | Set By User | Unit | Default |
|---|---|---|---|
com.unraveldata.monitoring.fs.partitions.check.interval Frequency to check monitored partitions. 0 disables monitoring. | s | 60 | |
com.unraveldata.monitoring.fs.partitions.csv Comma separated list of monitored partitions (named by folders). Symbolic links are supported. | string | /srv/unravel,/usr/local/unravell | |
com.unraveldata.monitoring.fs.partitions.high.watermark Trigger alert if disk usage is over this limit. Do not trigger next alert until disk usage is below low watermark limit. | percentage | 85 | |
com.unraveldata.monitoring.fs.partitions.low.watermark If disk usage must go below this limit before the disk alert can be triggered again. | percentage | 70 |
Deprecated Partition Properties
The following properties have been deprecated and replaced with the above partition properties. We strongly suggest you remove these deprecated properties from unravel.properties and replace them with the above properties.
Note
If you have both the new and deprecated properties defined, the value of the deprecated property is used. For instance, if you have set:
new: com.unraveldata.monitoring.fs.partitions.high.watermark=80
deprecated: com.unraveldata.kafka.monitor.disk.high.watermark=60
Unravel uses the deprecated property value, so triggering occurs when disk usage is > 60% not 80%
Property | Replaced by |
|---|---|
com.unraveldata.filesystem.volumes.csv | com.unraveldata.monitoring.fs.partitions.csv |
com.unraveldata.kafka.monitor.disk.high.watermark | com.unraveldata.monitoring.fs.partitions.high.watermark |
com.unraveldata.kafka.monitor.disk.low.watermark | com.unraveldata.monitoring.fs.partitions.low.watermark |
Property/Description | Set By User | Unit | Default |
|---|---|---|---|
com.unraveldata.monitoring.fs.folders.check.interval Frequency to check monitored folders (directories). 0 disables folders monitoring. | s | 0 | |
com.unraveldata.monitoring.fs.folder_limit.pairs.csv Comma separated list of monitored folders and their size limits. The folder must be fully qualified and may be a symbolic link. Size limit is a whole number with optional suffix of KB/MG/GB. Folder and its size limit is separated by a colon, e.g., /srv/unravel:100GB. If no size limit is specified, size is evaluated as bytes. Folder:limit pairs are separated by comma, /srv/unravel:100GB,/srv/unavel2:10GB. | string | /srv/unravel:100GB | |
com.unraveldata.monitoring.fs.folders.low.watermark Percentage of folder limit above. Once an alert is triggered, a new alert is only triggered if the size first drops below this limit and then rises above the folder limit. The purpose is to prevent false and repeated alerts. Example: /srv/unravel:100GB with a low water mark of 80% The first time (/srv/unravel size > 100GB) the alert is triggered. No new alert is triggered unless (/srv/unravel size drops < 80GB) and then (/srv/unravel grows > 100GB). | percentage | 80 |