Monitoring Setup
Configure how targets are monitored and how users are notified upon a failure.
The concept of Monitoring-Setup is to use the filter in order to apply settings to a single or multiple targets.
If you don't set a filter, all targets are updated at once.
Filter
- Use the templates (icons above filter) or click on the links of Target (to match a single target)
- Clicking on a test icon (e.g.
) executes a monitoring test on this target
- Clicking on Alert or Events Action (e.g.
) from the list applies it as filter
Monitor
- Define the Test
(Should be uptime for all switches and routers already)
- Setting it to "No" skips active polling. Can be used as maintenance mode or if you just want to set event-actions or discovery thresholds on a device
- Select icmp if TCP ping doesn't work on a target. Enter # of packets in
, if you want to send more than 1
- Test http/https: You can enter a string like "index.html" in and a regexp matching a successful response in
. Only a SYN check (TCP ping on port 80) is performed, if you don't
- Test dns: you can send a hostname and a regexp matching the expected IP address
- Test ntp: you can send RFC2030 fields like "Stratum" and enter a match ^[1-5]$ to detect if your ntp server lost sync
- Clicking "Update" applies the settings to the displayed targets
- Clicking "Delete" removes the displayed targets from monitoring
- Select email or SMS alerts, just have incidents create Monitoring-Events or nothing at all. If you select a repeat option, the alert is resent every 100th failed test
- The Latency textbox allows for changing the latency threshold for individual targets
- Click on
to simulate an outtage of the first monitored target
Events/Threshold
You can forward events as emails based on their level or contained text:
- With Forward in the first box select a minimum event level
- With Forward in the first box enter a regexp as the Filter
- Alternatively you can select Discard, a maximum event level and/or a regexp and matching events will not even be stored in the DB (Level limit can only be used to forward OR discard but not both)
- Setting a regexp for Maximum raises matching events to level 250 (Emergency) and shows those within the past 24h in Monitoring-Health (useful to identify failed power supplies or stack members)
- The notify settings from nedi.conf can be overridden for each target in the "Discover Notice" field
- To clear any fitler enter a "-" by itself
Reset
Sets dependency info, if available via links or device information (in case of node targets). After that, the dependencies can be adjusted on each target individually
Updates target IP address from devices or nodes (in case they've changed, there's a 
icon in the target status)
Reset the availability counters (lost & ok) once a year if you need to know annual availability for example
- A yellow/shaded target status indicates that its not found as node or device anymore (and should probably be deleted)
NeDi Help