Track Internet disconnections, provider outages with historical data, and automated speed testing.
For Windows, Linux, ARM64, ARMa7. Learn more by visiting www.outagesio.com
Notice: If you created an account on app.outagesio.com, simply use the same credentials to log in here.
Missing help text for outages
-
Hi,
Thank you for bringing this to our attention.
Strange, the share works in some tests but not in others.
The help text will also be fixed tomorrow.
I'll bring these up as issues tomorrow and we'll test.
In terms of the outages, I'll have to look at your agent. Do you have one or two routers internally, before the provider?
When there are multiple private IPs in the mix before and with the provider, the algorithm can sometimes be unsure of what is what, meaning, where your LAN ends and where your provider starts.
We do have a manual way of telling it which hops are which so I'll wait for your reply first.
-
Hi,
In terms of missing emails, yes, definitely, you should receive emails if you have them set in your notifications which I see you have.
I'll have someone email the address on your notifications to confirm you are getting them.
Update: Thanks for letting us know you received them. I do know that we had a problem last week where a service was accidentally shut down by a piece of code. it took a while for anyone to notice that but it was fixed right away. Was it last week?
In terms of setups, I'll try to explain what I mean as it might help someone else who comes across this.
The most standard to us usually means the providers router/modem that the customers devices are connected to.
The next most standard to us is the above but where the customer then has their own router or wifi/router connected to their providers router/modem.
Sometimes, the algorithm is not able to determine where the providers network starts because everything is using private IPs.
For example you might have a router internally that is then connected to a router that is then connected to your providers router/modem.
In this case, you might have a mix of 192.168.1.x and 192.168.0.x or 192.168.x.x and maybe the provider starts at 10.0.0.x.
Or, you might have a static so your first device could be 192.168.x.x, the next being the hop to your providers router/modem being a public IP then back to private IPs inside of the providers network.
The point is that if you know your network and you know that the classification is showing the wrong network (local network or provider) then we do have a fix for that which can be done for you. It is not something we offer to the general membership because many of our members are not necessarily technical users and we don't want anyone to set something that will give them the wrong results.
So in the end, in more complex networks, the person has to know this.
Having said all of the above, which is meant to help someone else that might come across this post, I do see that your network is very straight forward.
I do see that every single outage says it was x.x.40.1 which is the local network and just one outage which is beyond the provider and irrelevant.
Looking at your reports, it shows that you have something on your network that is disconnecting for around 9 seconds or so.
You mention that it is unlikely because you were using the Internet at the time so there is no denying that.
The agent cannot make up outages, it can only log and report them so there is something going on that is not clear which requires more investigation.
Sometimes, code needs human intervention for everything to come together but I'm not sure where to go with this since there is very limited data. The reports simply show what I mention above. Maybe we can keep looking at it using this thread and see if we can figure it out.
-
So, is it saying the device x.x.40.1, which is the router is going down for 9 seconds, or that something on the subnet is going down
for 9 seconds? I can see on my managed switch the x.x.40.1 device has an uptime of 6 days. What does outage.io consitute
an "outage" in this case?
The agent doesn't know if the router is going down or not, only that it cannot reach it for the amount of time it reports.
Even if a device is up constantly, it doesn't mean there could not be some other problem.
Maybe a cable is slowly going bad, maybe dirt/dust in the connector is causing a bit of loss, maybe something is hijacking the packets/path intermittently. The agent cannot know any of these things, it can only report what it experiences as just another device on the LAN.
We have seen plenty of cases where this kind of reporting didn't seem to make sense but we have to keep in mind that sometimes in networking, these are just the clues we are being given to dig deeper into the mystery. For example, as you said, you were on the connection at the time, would have noticed if it was being flaky and you are right but there does seem to be something happening. What it is will need more digging.
Let's look at what we've got.
Your outages show only one outage outside of your network and it is beyond your provider so it's irrelevant of course.
The rest of the outages are all on what appear to be your gateway at x.x.40.1.
Next, you can see which routes the packets have taken fairly currently and historically by looking at your Historical, Hops report.
Hop 2 which is your provider, has changed a couple of times which would not trigger an outage report.
An outage is recorded only and specifically if the agent cannot reach a hop in the path to the OutagesIO network. Meaning, it can only report an outage if there was a TCP/IP interruption and confirmation of this by the agent.
When the agent recognizes that it cannot reach the OutagesIO network, the first thing it does is to confirm that the problem isn't just something at the OutagesIO network side that's down. Meaning it runs a test to see if the service / server is responding and if it does in any way, then it means there is no outage since it can confirm that. If it cannot reach the OutagesIO network, it then tries to figure out what the last hops were that it was able to reach to determine which hop went down.
The fact that the agent is recording outages and not just showing Inactive and eventually Disconnected means it was able to determine a TCP/IP problem of some kind.
Is your router a router only or it is a wifi router?
It would be interesting if you had our hardware agent to specifically monitor the wifi provider.
Monitor your wireless, LTE, 4G, 5G, mobile Internet
I don't know, let's keep digging and see what we come up with.
-
By the way, in testing the share, for me, it didn't work using Chrome but worked for someone else here. It worked using firefox but I've not had anyone else test it yet.
It's now in our list of things to look at but we have an installer issue that is taking up most of our time right now.
-
There are no teething issues, just limitations in what software alone can do and we are constantly trying to improve as we come across new situations that help us to learn or think about new ideas. Sometimes, it takes human intervention to fully understand the situation.
I have had a constant ping going to the router from the agent computer and it hasn’t logged any dropped pings as of yet.
Unfortunately, I cannot tell you any more than I have since I do not have access to your network. We do come across situations that need human intervention and it seems this is one of those cases. Use the reports as leads and keep in mind that the agent is not going to do anything on its own, it's just software programmed to do specific things.
Why would it be able to monitor Wi-Fi though, does it have radios?
Yes, we use off the shelf devices that have WiFi built in and you can enable that from your control panel in Extended reports.
I would not suggest buying one since you are not 100% certain the service is working for your situation however or at least until you figure out what is going on.
As usual, we are happy to help but like you, we'll need more info to get to some point of understanding.
-
Well, in the last few days i have run into bugs with chrome not working
A reply was shared showing that the problem is not with the app, it's something to do with Chrome. I thought you were talking about an older problem that we used to have where sessions were lost. My mistake for not looking at the error in the image closely and jumping to that conclusion.
The problem is that most people do not report problems so they end up being there until someone does if we don't notice them. We practically beg in several places on the site asking that problems be reported so we can fix them but most still never do. No matter how much testing we do, sometimes we don't see them until someone posts to thank you for letting us know.
the graph images not downloading
As i said, this and the other items, the help buttons were going to be looked at today. In fact, I think the help buttons issue was already dealt with and the image download is being worked on right now.
the agent not reconnecting after a device reboot
I've not read that in any of your posts so this is new to me.
Speedtest hundreds of percent different then ones I run from other providers, etc.
Actually, fast.com is one of the speed test sites that the agent uses so what ever is causing the issue, it's not with the agent. In terms of your connection, I've also pointed out that it is very odd, it's all over the place. I shared what I saw in one of the replies.
difficult to determine if an outrage it does report is perhaps just another bug.
The reports are correct. As explained in another post, if you are seeing outages, this is TCP/IP based and the agent cannot make these things up. It can only report what it sees.
You have opened many posts now and I am having a hard time keeping up. We are a small team that works non stop on this project, we are not investor funded and want to offer a service that can help the public with a problem that causes frustrations to a lot of people.
I am happy to give you your money back for your Extended level if you feel the service will not work for you but as you can see in many other posts, it works fine for many. As I've also explained, sometimes and especially in this case, human intervention is required because what ever is happening with your connection, I do not think that automated software can figure it out.
-
We've fixed the help text problem and I believe the download image problem was also fixed.
The problem with the browser is not something on our end. I shared a link showing what the problem is.
We always consider that problems are possible and hope that members report them as you have so we thank you for that.
We've fixed it, can you check.
