OutagesIO Support ForumsInformation, Support and FAQ relating to OutagesIO. Monitor your Internet service and provider with optional alerts to connection issues or downtime. For home, small business, IT companies and enterprise. Tracks Internet and provider reliability with useful proof. For Windows 7, 8, 10, Centos 7, 8, Debian 7, Ubuntu, ARM (Raspberry, Tinker Board, etc).
Learn more by visiting our web site, www.outagesio.com
outages.io in a docker container
I want to run outages.io in a container and then add it to the Unraid appstore.
I have done this using Debian but it's quite large, also the agent isn't connecting. I'll post about that after I've done some more troubleshooting. Links below for reference.
Ultimately I'd like to run this on Alpine Linux so it ends up small, but I get the following issue that prevents me from running the binary, any idea what this issue is?
/otm # ldd otm_linux
libpthread.so.0 => /lib64/ld-linux-x86-64.so.2 (0x7f360a322000)
libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0x7f360a181000)
libm.so.6 => /lib64/ld-linux-x86-64.so.2 (0x7f360a322000)
libgcc_s.so.1 => /usr/lib/libgcc_s.so.1 (0x7f360a167000)
libc.so.6 => /lib64/ld-linux-x86-64.so.2 (0x7f360a322000)
Error relocating otm_linux: __res_init: symbol not found
Hi, running the same against our Linux version on Centos 8, I get the following.
# ldd otm_1.67.2104_linux linux-vdso.so.1 (0x00007ffe0fb6b000) libpthread.so.0 => /lib64/libpthread.so.0 (0x00007f47a4aa0000) libstdc++.so.6 => /lib64/libstdc++.so.6 (0x00007f47a470b000) libm.so.6 => /lib64/libm.so.6 (0x00007f47a4389000) libgcc_s.so.1 => /lib64/libgcc_s.so.1 (0x00007f47a4171000) libc.so.6 => /lib64/libc.so.6 (0x00007f47a3dac000) /lib64/ld-linux-x86-64.so.2 (0x00007f47a4cc0000)
I suspect it is because our Linux version was built on Centos 7.6.1810.
On our Aarch64 version;
# ldd otm_aarch64 linux-vdso.so.1 (0x0000007faf011000) libpthread.so.0 => /usr/lib/aarch64-linux-gnu/libpthread.so.0 (0x0000007faef73000) libstdc++.so.6 => /usr/lib/aarch64-linux-gnu/libstdc++.so.6 (0x0000007faed8e000) libm.so.6 => /usr/lib/aarch64-linux-gnu/libm.so.6 (0x0000007faece1000) libgcc_s.so.1 => /usr/lib/aarch64-linux-gnu/libgcc_s.so.1 (0x0000007faecbd000) libc.so.6 => /usr/lib/aarch64-linux-gnu/libc.so.6 (0x0000007faeb4a000) /lib/ld-linux-aarch64.so.1 (0x0000007faefe1000)
Unfortunately, our dev that works on the agents has been sick for a while, we are unable to provide much more information right now.
I don'k know if there is a work around for something like this.
On the other issue, can you start a new post when you're ready with that test and we'll certainly try to help.
Have you had any luck with this?
Were you able to solve the other problem that you mentioned for another post?
I didn't change anything my end, reinstalled my container (which pulls down the latest agent) and I think it's connected now. I'm not able to see any stats though because seems you charge for this now. How do I verify it's connected?
@peachy Let's see what's happening. Can you share the agent ID please.
@outagesio_support sure, 128433
Yes, there is a small subscription fee for Extended reports. Development hours on this project are insane and costly.
I've upgraded that agent to help in your development effort so please don't delete that agent :).
You'll see it's sending pings and if you enable speed testing, you should see that too. Once you have outages, you'll see those.
Can you expand on how to fixed the problem 'linux-vdso.so.1'.
I haven't fixed it, the Debian based container works now so I guess you updated the agent in the last few weeks. The issue you refer to occurs in the alpine Linux container, I could take a look at that later.
So I think the docker image I provided works so you can probably run that at your end and test it. I also provided a template to run this in unraid which should work on unraid too.
I do think that the data provided without subscription isn't enough for people to feel it worth to run the container I created, personally if I couldn't verify it was running I'd just delete it. Happy to discuss this further of you want to.
We've tried a number of different ways to have a free version and earn enough to keep the service going.
We are always open to ideas but developers and resources aren't free, all of it is expensive and incredibly time consuming so we need to bring in some income.
We do keep it inexpensive as possible and the free version seems to work for most people which is how we came up with the pricing.
Always open to ideas of course.
@outagesio_support I guess I'd look at it the other way and think of who would pay for the paid version then try to increase userbase on the free. Users like myself would never pay the fee for premium not because I think the service isn't worth it (i'm here discussing with you after all) but because for $70 a year i could get a small vm for the same sort of money I could also do other stuff with.
The data produced has value too because if it gained enough support it could hold ISPs to account, which is what attracted me to what you guys were doing in the first place.
Where are the costs? My guess
- Development which you have to do anyway
- Hosting the service (graphs, analytics)
- Hosting the endpoint
- Building support for the agents on different platforms.
The last two you could get community support with, the first two if you wanted community support you'd have to open source your software which you likely don't want to do or you would have done it already. Maybe if the agent wrote metrics locally in a logfile or something that could be included in free?
I guess I'd look at it the other way and think of who would
pay for the paid version
Just to jump to the main point... we do aspire to make the service free at some point.
In the meantime, that's exactly how we got here after many years of running this and trying a lot of variations.
It was completely free and the more agents that got installed, the more costly it became not only in terms of resources and management of that but in terms of development.
We learned that even free, it was amazingly difficult to market. Since adding paid options, we've also learned that consumers don't want to pay for a service that helps them to fix another paid service.
We offer the free version because it helps just about anyone and without overwhelming details. The average user is not interested in all the details, they just want to know what's happening to their service.
However, some folks do want to more details and for them, it's worth the small cost to upgrade. They can also let the upgrade lapse once they solve the problem while keeping the agent running so as to have data when things get bad again.
then try to increase userbase on the free.
This is not an easy task. Everything is against us when it comes to marketing. We have to try and be seen against large corporations that are willing to spend literally anything to get the clicks.
Keeping in mind that our competitors are large providers, big outage sites, speed test sites, all big dollar entities that would prefer to get your attention to their outages sites rather than actually finding out what's going on in your own area where most of the problems really are.
It is extremely expensive to advertise and try to get a tiny result because they all use the same key words / phrases that we use. SEO is a massive pain, it's a non stop moving target as Google and others keep changing the rules almost daily.
What makes it even harder is that consumers could care less about leaving reviews for example which is something that Google and others practically demand these days. If you don't have any reviews, you don't get found easily.
We have a hard time being found on the Internet and that doesn't help anything.
Users like myself would never pay the fee for premium not >because I think the service isn't worth it (i'm here discussing >with you after all) but because for $70 a year i could get a >small vm for the same sort of money I could also do other >stuff with.
We want to find a balance where we can keep a free version going while earning our keep at an affordable price.
The service was not built for developers, it was built for consumers that aren't very technical while having a side that is highly usable for IT support.
The data produced has value too because if it gained >enough support it could hold ISPs to account, which is what >attracted me to what you guys were doing in the first place.
There are a number of accounts in the forums and in our blog from members that were able to hold their ISP accountable using our free version.
That is the main goal of the service, giving people an easy to use tool that gives them the proof they need to get help from their provider. Some people go through hell having to fight their provider to get them to fix anything. This service is for them.
Where are the costs? My guess
Not something I'm going to get into because everyone always feels it's easy at the other end, doesn't take much, etc.
Suffice it to say it's cost us both in terms of life savings and then some since we are a bootstrap and in terms of life quality itself. It's a beast that demands a lot of time, a lot of people and is constantly in flux. There are a lot of moving parts.
Development which you have to do anyway
If we didn't work on the service, then it's not something we would be doing anyway :).
Building support for the agents on different platforms.
This is another tough one because OS's are constantly changing. It's almost a full time job working on not only the software development but sourcing the hardware. The majority of the hours are spent troubleshooting problems.
The last two you could get community support with, the first >two if you wanted community support you'd have to open >source your software which you likely don't want to do or >you would have done it already.
This is not an easy one either. As soon as you open code up, you open the door to losing everything you've put into something.
In open source, everyone wants everything to be free and to be shared. People are happy to take your code without one thought about what went into that to get it here.
We would love to offer the agent code but we've not found a way that would not compromise everything we've put into this for years now.
Since we've come out, we've found plenty of people hacking away at the agent and the site and now see copy cat services.
I'm sure there will be free versions until they run out of steam, get taken over to the point where one large company owns it all with few choices and lots of patents, regulations, who knows. That's how things always seem to go :).
Maybe if the agent wrote metrics locally in a logfile or >something that could be included in free?
How could we earn anything to stay in business then? People would simply nab the agent, write code, make it available to anyone that wants to download it. In turn, that would generate huge amounts of traffic and resources being used while earning nothing what so ever.
There are some serious costs behind all this and there needs to be some income. The free version seems to work well for many so maybe the focus should not be on free but on value.
Everything mentioned above are not free and the service doesn't run on simply a few VPS.
We would love to see an open source aspect of this service but we're just not there yet. The costs to get to where we are now have been high and the work load never ends.
Trying to figure out simply how to be found on the Internet continues to be our biggest challenge. Once we can figure those things out, all kinds of interesting things could happen including offering more data at the free level.
We do believe it would be nice to work with others that could access the agent source code but we've not found a way of doing this in a controllable way.
We thought we might start with something like this.
Internet of things data in your OutagesIO dashboard
The idea being that perhaps it could be a way to start working with developers that could lead to more. One step at a time.
i think if it got into the Unraid app store you'd probably get a lot of installs, but I think you'd at least need the heartbeat status included in the free version otherwise it would be unsupportable.
Free includes everything in the light view right?
Anything is doable. We do have something very interesting coming that should help simplify everything but we've been having a hard time finding a developer that also has the mind of an artist for the graph we need :).
We could continue this in emails and see if we can get to something.
sure let's do that