Resolved Monitoring Plesk port via uptimerobot.com makes trouble since (ca.) version 18.0.65

[King555]

Server operating system version

Ubuntu 18.04.6 LTS x64 (ELS)

Plesk version and microupdate number

Plesk Obsidian 18.0.65 Update #1 Web Admin Edition

Is there anybody here who monitors his Plesk installation via uptimerobot.com? Since the 18.0.65 update, but maybe shortly before that, UptimeRobot constantly tells me that the port is not reachable, then after an hour, a day or so the port is reachable again, and so on. I monitor only the port, not the HTTP(S) site.

Maybe this it totally unrelated to Plesk, but I don't know. It almost never happened before. There was a time when the port was completely unreachable for some time for UptimeRobot (from 10/2023 to 05/2024, initiated and solved by Plesk updates), but the difference now is that it's sometimes reachable.

This monitoring worked from 09/2018 until today except for the timeframe mentioned. The Plesk installation in my browser (same port, of course) works always, I never had a problem and also it works at the moment when I get a notification from UptimeRobot that the port is down.

Any ideas? Is there some kind of protection in Plesk? Or do I have to search the solution on my server's side? UptimeRobot sends me many mails and it's annoying (up, down, up, down). Of course I can disable the mails, but I want to monitor Plesk.

 

[Martin Dias]

Locations and IPs | UptimeRobot

If you get any false positives, there is a strong chance that the IPs used are blocked by your hosting provider. Make sure that you allow-list these IPs!

uptimerobot.com

Check if their IPs are landing in Fail2Ban for example

 

[King555]

None of their IPs are blocked by fail2ban. The firewall (iptables) is static, so that's also not the reason. ModSecurity is just for HTTP, if I'm correct. I think that's all I have installed for protection.

UptimeRobot can monitor everything else on the server without any problems: multiple websites via HTTP, several services via the corresponding ports (for example all mail ports) and also a ping of IPv4 and IPv6 addresses. It's only the Plesk port which makes problems.

 

[King555]

I tried to find out whether there is a pattern when the port is not reachable, but I can't find one. Except the time frame in which the port is available is always 6 to 7 minutes. The following list contains times, every pair contains first the up time and then the down time (24-hour format). These are different dates, by the way.

00:39
00:45

09:39
09:45

11:28
11:35

11:55
12:01

00:43
00:50

12:16
12:22

04:32
04:39

06:22
06:29

07:41
07:48

01:54
02:01

I'm giving it up for the moment, I disabled monitoring of the Plesk port completely now.

 

[King555]

Since I switched to a new server with a fresh Plesk installation and Ubuntu 24 instead of 18 about a week ago, monitoring works again. Let's see how long, but at the moment this topic is solved.

 

[User15]

Hey @King555 ,

I have the same issues with my plesk servers and uptimerobot.
They are running e.g. on Ubuntu 22 - could you find a reason for the issues - or is still everything solved since your update to Ubuntu 24?

 

[King555]

I can only say that it works since about 10 days. Maybe the problem comes back, I don't know.

Unfortunately I did not find out what the reason was. But now everything is new; new server, new IP, much newer and freshly installed operating system and fresh Plesk installation.

 

[Kaspar]

Hey @King555 ,

I have the same issues with my plesk servers and uptimerobot.
They are running e.g. on Ubuntu 22 - could you find a reason for the issues - or is still everything solved since your update to Ubuntu 24?

Did you (double) check if any of the IP addresses used by Uptimerobot for monitoring are being banned by fail2ban on your server?

 

[User15]

Did you (double) check if any of the IP addresses used by Uptimerobot for monitoring are being banned by fail2ban on your server?

Yes, I checked all of them - nothing found.

 

[King555]

Since the update to 18.0.67 it does not work again. For me it looks like an issue with Plesk, not with Uptimerobot. But I'm giving it up now, because all other monitors still work fine and should be sufficient for monitoring the server.

 

[User15]

I'm also still having issues - 177 incidents in the last 24 hours // and Plesk Support also double-checked and confirmed that UptimeRobot IPs aren't in my fail2ban logs..

 

[User15]

Feedback from UptimeRobot Support is that it probably is an issue on their side due to issues with old IPs.

Fix would be: Delete monitor, Create a new monitor → then UptimeRobot will automatically use more recent IPs without issues
An automated change of all existing monitors may come at any time in the future, perhaps, …

 

[King555]

I will try that, but I'm pretty sure I already recreated the monitor while having this issue.

 

[King555]

It works now since 6 days.