Every OLT, switch and router in your rack is watched for uptime, for the state of each port and for its own health. Switch on SNMP read-only and HansaNet works with the box as it is.
Uptime first
Every device you enter, OLT, switch, router, CPE or NVR, is pinged by the probe in your rack. A real outage is confirmed in about a minute, a single dropped ping is filtered, and the devices fed by a failed one are held back so one cut is not a storm of pages. That part is outage detection.
For a tech standing at a device there is live ping: the results from the probe to that one device on the screen, one line per packet. A second viewer joins the same session rather than adding traffic to the box. Workers can open it, from the phone as well. It writes nothing, so watching a device pages nobody.
Ports: what is up, and what is moving
Each port shows whether it is up, whether somebody has switched it off, and the traffic going each way. Names, aliases and speeds follow the ones set on the box. A port that goes down becomes an issue on the issues page and is counted on Home.
Small OLT and switch agents are slow and often answer only part way. HansaNet works with them as they are, and you do not have to change the box. When a read comes back incomplete the ports panel says so, instead of showing a number it does not have.
Health, and the Test SNMP verdict
CPU, memory and temperature come through on the vendors we support, and the OLT pages say what you get on each family. Adding a vendor is a change on our side. Nothing is reinstalled in your rack.
When a box will not answer, an owner presses Test SNMP. It says whether the device answered and what it calls itself if it did. If it did not, it names the fault: the community string, an ACL, or UDP 161 blocked. Every one of those is on your side and yours to fix.
A monitored device is Rs 100 a month minimum. Passive plant is never counted. See pricing.
What it does not do
- It watches shared infrastructure, not subscribers' home routers. Per-ONU readings are on optical monitoring.
- Bandwidth is an average over the last few minutes, not a per-second graph.
- A renamed port takes a little while to show its new name.
- A read that came back incomplete stays incomplete until the next one.
- Health readings need a vendor we have added. A vendor we have not met yet gets uptime and ports.
- ISP accounts get the Test SNMP verdict, never raw device values.
Related: outage detection, the device web UI, uptime reports.
Put a probe on your plant tonight.
One setup .exe on a Windows machine in the rack, or a .deb on Linux. Rs 3 per ONU a month, no plans, no caps.