Agent Ping graph stopped working

[RobotWookie]

I have my agent set up and working. Last night it was recording pings and showing them on the graph in the dashboard. When I checked it this morning, the heartbeat/status graph is still working, but pings is empty. I also checked the historical ping graph and it was blank as well. Any ideas why the graph would stop showing, while everything is still up and running?

I'm running the agent on a Raspberry Pi with an ethernet connection.

[OutagesIO_Support]

Hi,

Can you share the agent ID so I can take a look at that to see if I notice anything unusual first.

[RobotWookie]

@OutagesIO_Support The agent id is 129221. Thanks!

[OutagesIO_Support]

Hi,

Well, for what ever reason, it seems like your agent or network has blocked the ICMP based heartbeat tests somewhere, somehow.

I looked at many other Pi devices running the same agent version and didn't find any that we not using ICMP.

Is it possible you have something that might decide to block traffic after a certain amount has passed, like an auto blocking rule or something that is not aware of this traffic being allowed?

What ever the problem is, it must be on the Pi , the local network or even the provider possibly auto blocking.

Can you ping tpw.outagesio.com for example from the same Pi device?

[RobotWookie]

@OutagesIO_Support I am able to ping tpw.outagesio.com from the Rpi. I don’t know of any settings on my router that would be blocking traffic. But any help in pointing me in the right direction would be much appreciated. I have been having major issues with my ISP dropping service so I’m trying to collect data to present to them.

Thanks!

[OutagesIO_Support]

@RobotWookie said in Agent Ping graph stopped working:

129221

Did you install the agent as a service or just running it manually using the script? Can you kill the agent process, restart it and see if pings come in? It stopped yesterday for some reason.

[RobotWookie]

@OutagesIO_Support I am running it manually at the moment. Stopping it and restarting it does seem to have brought back the pings. I guess we will see if they are still coming in tomorrow morning. Also, can you point me to the page with instructions to set it up as a service. I saw the link during the setup process, but I was more concerned with getting it working at all so I didn't bother with that step. But now I can't find the page again.

Thanks!

[OutagesIO_Support]

Ok, so at least we know that pings can be sent when the agent is able to get them.

When I checked earlier, the agent had no pings to send which means it wasn't getting any results from the operating system.

The pings aren't really important in terms of being an accurate measurement. They are mainly being sent as information that could be useful in some situations but mostly to see if there is a substantial change in latency between point A and point B.

If pings latency goes up by a certain percentage, the agents algorithm considers that this could mean something happening and can trigger a speed test for example and other tests.

However, if pings aren't being sent, it means the agent is not running right as ICMP is one of several functions used to also determine if there is an IP outages.

All that to say, it's important that you see pings being sent to the dashboard.

Since restarting the agent got pings going again, to me, it means maybe there is something on the OS that is monitoring certain behavior and limiting something like ICMP.

I would rule out the agent as having a bug because we would have seen it by now as there are many others running the same version.

These can be a bit hard to figure out sometimes. Let's see what happens next.

In terms of running as a service, I think this is what you are looking for.
Monitor your Internet with raspberry pi, nanopi and other arm devices