Outages not detected

[pnagl]

I just installed my first agent to troubleshoot a home internet problem.

I had several internet outages today but none of them were recorded as outages on the dashboard so the details were lost after an hour. What defines an outage.

the outages were captured in my heartbeat stream.

Each heartbeat dropout was about 2 minutes.

10:04:30 to 10:06:23

10:13:20 to 10:15:10

[OutagesIO_Support]

Hi,

Welcome to OutagesIO.

This might help explain what the agent considers outages.

About: Internet outages, alerts and agent statuses

If you're not too familiar with how IP works, this could help also.

The nature of TCP/IP and Internet connectivity

Hope this helps and if you can post the agent Id, I'll take a closer look.

[pnagl]

I see where it talks about if you "do not see outages in your reports, it could mean a hardware/signal problem somewhere"

So that must extend to any issues effecting an individual residence - right? So an outage must then be something that effects multiple customers for a provider. How does the app know when its an actual outage?

If I want to see when I have outages even if they effect just me do I need to either watch the heartbeat for blips or better yet turn on the email alerts to know when the heartbeat issues happen? Is there a better way?

Thanks

[OutagesIO_Support]

it could mean a hardware/signal problem somewhere"

So that must extend to any issues effecting an individual residence - right?

Yes, correct. It could be the modem/WiFi device at the customer location and/or it could be with the provider. There is no way to know when it comes to signal problems because only the provider controls their own network and they do not share this kind of information.

So an outage must then be something that effects multiple customers for a provider.

Not always but if you were to eliminate signal levels at your location and still see either no connectivity or outages, then chances are others in your area are also affected. If you were to motivate one or two others in your area to use OutagesIO, then you could more easily confirm these things.

For example, if you are watching the map at the time of a problem, you would see other agents in your area also going into an 'Inactive' state.

If there was an outage then it is likely that your location and your neighbors would show a matching outage and/or date/time, hop, IP, even the provider name if you are all using the same provider.

How does the app know when its an actual outage?

The app part are the reports that you're looking at when you log into app.outagesio.com. The agent is what is monitoring the Internet connection. It is purely IP based monitoring. In the simplest description, it maintains a constant ongoing connection to one of our networks using standard port 80.

If something becomes inaccessible from source (the agent location) to destination (one of our networks), it immediately starts running a series of IP based tests to confirm that there is a hop down somewhere.

When you see Inactive, it does not mean an outage, it is simply a notice that your agent is no longer able to communicate with our network. IF there is an outage and you did not restart the agent, it will send that report once it is able to communicate again. If you keep seeing Inactive but you are not seeing outages, it means the problem is not purely IP based, it could be any number of things. Knowing a bit about the kind of network the provider is using, coax, DSL, Wisp can help to narrow the possibilities. Getting others in your area to also monitor can give a much clearer picture of what is going on in your area.

If I want to see when I have outages even if they effect just me do I need to either watch the heartbeat for blips or

better yet turn on the email alerts to know when the heartbeat issues happen? Is there a better way?

The heartbeat does nothing but let you know the agent is communicating. We just wanted something easy to visualize so that members could clearly tell when their agent was communicating. The heartbeat simply means that the agent is communicating every 15/20 seconds or so. You cannot determine anything by the heartbeat other than the agent is currently communicating or not.

In the basic reports, the options are to receive an email for any Inactive condition. It is just a way to get your attention so that you can come to the dashboard and see what is going on. Again, if you see inactives without outages, you are likely not experiencing outages.

Having said that, I am talking about the majority of consumers that have fairly standard setup, not multiple switches, multiple routers and other things that could throw the algorithm off. In most cases, it's pretty accurate.

In your case,it seems that the provider controls your modem, changing its private IP fairly regularly. Maybe that is indicative of some problems in your area too.

I hope this helps you, it's a lot of information :).