Anyone else use Monitor.us? How many false positives do you receive?

[SGC-Hosting]

I'm continuing to use monitor.us alongside several other monitoring services - but the number of false positives from monitor.us is just ridiculous.  Anyone else experiencing the same issues?

I have 9 servers monitored by them -- and if I were to follow their uptime reports, I'd have below 98% uptime.  Whereas - the other platforms I use are finding 100% uptime (also confirmed with pings/logins from my phone and pc after receiving alerts).

Anyone else experiencing similar issues?  This goes for ping tests and specific service checks.

While we're at it - what service do you use and what's your experience?  I signed up for the free service from statuscake to test it out and it has been great so far (no false positives)-- I haven't had a chance to test out experience with a service/server that has gone down though.  My only gripe is that I liked the monitor.us interface a lot better.

 

[DomainBop]

I'm using Monitor.us to monitor 57 devices and I've found them to be fairly accurate during the 2 years I've been using them. 

The "false positives" you're seeing with them (and that you'll occasionally see with any offsite monitoring service) are usually due to  temporary networking/routing problems between the test servers and your servers so your server may be up and reachable by the majority of people but the odds are it is temporarily unreachable to some people on some routes/locations.

I also use Observium, Zabbix, and StatusCake (paid) for offsite monitoring (as well as webmin's monitoring module and a couple of custom scripts for onsite monitoring and restarting services). I'm using 16 VPSs to monitor from every continent but Antarctica...I like to stay informed

 

[Francisco]

We use nodeping and it's iffy.

We've had LOTS of false positives where it'll email us that a whole location is offline (or at the very least the core router), all the while i'm logged into said servers over SSH, loads are low and couldn't be happier.

Their SSH monitors seem to be the most gimped of the bunch. I'm assuming they have fewer locations to monitor SSH or they are getting blocked by some LEB's crappy IDS.

Francisco

 

[SGC-Hosting]

I'm using Monitor.us to monitor 57 devices and I've found them to be fairly accurate during the 2 years I've been using them. 

The "false positives" you're seeing with them (and that you'll occasionally see with any offsite monitoring service) are usually due to  temporary networking/routing problems between the test servers and your servers so your server may be up and reachable by the majority of people but the odds are it is temporarily unreachable to some people on some routes/locations.

That was I thought at first.. but every day it becomes more and more frequent -- I had over 20 "problem" alerts on one server alone from monitor.us today yet 0 from all the others.  Everything was fine up until about 2 or 3 months ago.

What locations are you using to monitor?  My errors are about 50% hostname unknown and 50% 4 of 5 packets lost.

 

[DomainBop]

What locations are you using to monitor?  My errors are about 50% hostname unknown and 50% 4 of 5 packets lost.

I'm using their Germany and US-mid locations but most of my servers (and 100% of my websites) are in Europe so the results I see from their monitoring will probably be different than someone who is monitoring primarily US-based servers.

50% hostname unknown

That seems like they're having trouble communicating with your DNS servers.

We've had LOTS of false positives where it'll email us that a whole location is offline (or at the very least the core router), all the while i'm logged into said servers over SSH,

The day isn't even half over and I've already had a couple of those.  About an hour ago my Observium install in Milan alerted me that all my servers at OVH were down (both France and Canada servers).  False alarm, temporary routing problem in Milan...

traceroute to [server at ovh]

2  gw-cdlan-2.prometeus.cdlan.net (217.171.46.253)  0.651 ms  0.823 ms  1.002 ms


 3  ibgp-gw-core-a.cdlan.net (217.171.32.129)  0.706 ms  0.888 ms  0.993 ms


 4  * * *


 5  * * *


 6  * * *


 7  * * *

 

[SGC-Hosting]

That seems like they're having trouble communicating with your DNS servers.

Yep - that's what I thought, but I have not been able to confirm any issues with the relevant DNS servers for those machines.  The only time I was able to find an issue was a DNS server was actually down.

edit:

I'm using the same test locations and the servers are scattered across the US.  I guess I'll look into the issue further.  Thanks.

 

[iWF-Jacob]

We use StatusCake and it's been continually going downhill -- we have false positives all the time ranging from a couple seconds to two minutes, which is really quite annoying. We used to use Hyperspin which was pretty good -- never had a false positive there!

 

[Schultz]

uptimerobot also has false postives, from 1-5 minutes at random.

 

[SGC-Hosting]

We use StatusCake and it's been continually going downhill -- we have false positives all the time ranging from a couple seconds to two minutes, which is really quite annoying. We used to use Hyperspin which was pretty good -- never had a false positive there!

I'll have to add hyperspin to the list -- i remember coming across it a while ago, but never signed up and forgot about it.

Yesterday I added a cname to some of the servers with the most false positives (hostname1.domain.com cname -> hostname.domain.com)  "hostname1" was reported to have ~95% uptime while "hostname" had 91% in the same time frame (same monitoring settings obviously) while statuscake and others reported 100%.

It's frustrating because I want to be able to respond to downtime immediately and not have to wait for an alternative monitor to report it as well!  Oh well, such is life - I'll keep trying new systems.

 

[WSWD]

I've had horrendous experience with StatusCake as well.  You can see my post(s) on WHT about that one.  You think you have problems with false positives now.....

Went with NodePIng some time ago and haven't looked back.  They have been phenomenal. 

 

[dave]

I've used the free version of monitor.us for many years.  I haven't really noticed false positives, but there was a time it seemed like they were using stale dns entries.

 

[Tsanten]

I've used the free version of monitor.us for many years.  I haven't really noticed false positives, but there was a time it seemed like they were using stale dns entries.

Same here and from my experience the free version (30 min ) is better than Uptimerobot.

A week ago a webserver wen't down at 6AM Monitor.us send the mail at 6,30 and Uptimerobot at 8.00.

It's a rare case and for that reason i keep both, one for reliability and  another for the quick check.

 

[SGC-Hosting]

I now have quite a few monitoring services going and Monitor.us is still the only one reporting downtimes -- however they have reduces drastically the past few days.

I'll have to create a script that performs checks whenever I receive an alert from any of them... it will save me a bit of stress!