VPS server monitor

[allu62]

Hello after a very long time. My website running on a VPS since this bizarre affair with CPanel, I rarely visit the HelioHost forum.

Some days ago, two of my servers were down. As I haven't an Internet connection of my own, I very rarely check their status and I noticed it just by chance. My VPS running on Ubuntu server with Hestia CP, someone has an idea if there is a very simple application, that regularly checks all servers and when one of them is offline, sends me an email? Or, do you think that the best way to do this is to write a basic Perl script myself?

I have another question: At the beginning, I got several times a mail that Hestia has been updated to a new version. I haven't got any for a very long time and I wonder if there weren't any updates, if I just wasn't notified, or maybe there's something wrong in my configuration, so that there aren't any updates done anymore.

Thanks in advance for any help. And a Merry Christmas to everyone.

 

[wolstech]

Uptime robot is the monitoring service that I use. It's free for basic monitoring and can be set to send an email whenever something it monitors is down. It can monitor websites or ports, so you can point it at your website or at your SSH port.

As for hestia not updating anymore, that's a Krydos question...

[Kairion]

Latest version (1.6.14) of HestiaCP was released on 2022/12/11 (Y/M/D). Go to https://your_vps_address:8083/list/server/ and check which version you're running.

If it's below 1.6.14 then something went wrong with update. However if it's updated then it could be a problem with your mailing or HestiaCP isn't configured to mail you when an update is installed.

[Krydos]

Sometimes the Hestia updates break. I've seen it a few times where the main Hestia process refuses to restart for some reason, and the updates stop happening because they can't restart that process, but everything else continues running fine apart from that. A full reboot of the server might get the updates going again, but if the Hestia process refuses to start in addition to refusing to restart you might reboot the server and have even more problems. Just let us know and we can take a look at it if you need help.

The other thing that you might be interested in is a free rebuild. It looks like you're running Ubuntu 20 still. Ubuntu 22 came out 8 months ago. We've had a lot of people run into problems trying to upgrade from 20 to 22 so what we recommend is backing everything up, and then we can wipe it and install Ubuntu 22 fresh, and you can restore your backups. After VPS have been running for over a year like yours they tend to get little bugs and problems start to accumulate.

[allu62]

I hadn't yet the possibility to try Uptime robot, nor to check for Hestia CP.  Concerning Ubuntu 22, I think that a Linux upgrade with Hestia installed would be a really bad idea. And a fresh install? Backup of everything, install of Ubuntu, re-install and re-config of Hestia, restore of website, databases, scripts - that seems a lot of work to me. Do you really think that this is necessary? My VPS runs for a long time, but except this 2 servers down case, there wasn't ever a problem. Wouldn't it be more logical to follow the "Never change a running system" rule? I don't think that Ubuntu 20 has any security issues and as long as everything works fine, why taking all this necessary time to change (and even not being sure, if on the new system everything will work).

It not really belongs in this post, but it's VPS related. On December 26, I had over 6000 visitors accessing a Perl script, that reads a huge amount of data from a MariaDB database, sorts it and displays a list. The application, the script belongs to, is more or less paused during 2 years and normally there is just a handfull of people accessing this URI. I first thought that it might be some trial to overload the servers, but awstats reports the visits as "unique visitors", so all different IP addresses. And over 800 people, who were more than half an hour on my site. Does someone have an explanation of this? Is it possible to fake the IP? Could there have been something wrong with awstats? Or just 6000 people, having the same day nothing better to do than run this script?

 

[Krydos]

No, you definitely don't need to rebuild if you don't want to. I was just offering you a free service that you may not have been aware of.

Yes, it is possible to spoof an IP address as long as you don't care whether you receive the results to the webpage you access.

It's also possible your script was posted on a website or something and a lot of people read the post and tried it out. That sort of thing happens sometimes. A few years back a user on our free hosting made it to the front page of 9gag.com and literally hundreds of thousands of people were trying to access his website at the same time. We had to suspend his account and remove the A record from his DNS records to keep everyone else's website online. They call that sort of thing the "hug of death" on the internet.

[infantex]

Hello. 

I tried to use Uptime Robot to monitor the email service on Plesk for my two domains (infantex.com.mx and zaldivar.mx) but it reports them down.

I used port monitoring, that, supposedly, is the way to go, according to their FAQs. The only options are port 25 for SMTP and 110 for POP3 (and 143 for IMAP which I don't use), though.

Any recommendations?

Edit: I found out it's also possible to monitor "custom" ports. I set the monitors to ports 465 (SMTP) and 995 (POP). Hopefully this will work. I will post my results.

Edit 2: No luck.

Custom port monitoring also results in a site down status.

Edited January 12, 2023 by infantex
Trying custom ports option

[Kairion]

On 1/12/2023 at 5:07 PM, infantex said:

Hello. 

I tried to use Uptime Robot to monitor the email service on Plesk for my two domains (infantex.com.mx and zaldivar.mx) but it reports them down.

I used port monitoring, that, supposedly, is the way to go, according to their FAQs. The only options are port 25 for SMTP and 110 for POP3 (and 143 for IMAP which I don't use), though.

Any recommendations?

Edit: I found out it's also possible to monitor "custom" ports. I set the monitors to ports 465 (SMTP) and 995 (POP). Hopefully this will work. I will post my results.

Edit 2: No luck.

Custom port monitoring also results in a site down status.

Could you share what settings you're using on Uptime Robot?

[wolstech]

Uptime Robot is incompatible with Cloudflare when the proxy feature is enabled. If he turns off the proxying on CF for the root domain's A record, it will work. The problem he's seeing is because he's trying to check a nonexistent mail server in Cloudflare's data center instead of the actual mail server on Tommy. Similarly, a monitor for HTTP will have the opposite problem (always reports as up even when its not because CF will respond even when Tommy is down).

If you want to use the CF proxy feature, set Uptime Robot to watch the server's host name instead of your own domain (it'd work if he points it to tommy2.heliohost.org instead).

I use the same config and chose the tommy2.heliohost.org option since I wanted the proxy for speed.

[Kairion]

I had no idea he was behind Cloudflare 😅.

But you're right, I use Uptime Robot to check my VPS IP address on SSH port and it works fine.

PS.: I do have a HTTP check on my Tommy account behind Cloudflare and it does report when my site goes down. Right now it says my Tommy account has an uptime of 99,57%.

[wolstech]

Web monitor really depends. If your just watching port 80 with a port monitor, it will give bogus results of always up when used with CF.

If you do actual http monitor (which will make an http request and look at response code), yes it works on CF (CF returns a 5xx code when the origin server is down). 

[allu62]

So, Uptime Robot works well? Maybe I'll try it... But can it really monitor all services? I mean all those listed on the Hestia page (including iptables, spamassassin). Do these services all run on a given port and how to do to find out what port it is (sorry, my admin knowledge is not lots more than zero)?

On 12/23/2022 at 9:04 PM, Kairion said:

Latest version (1.6.14) of HestiaCP was released on 2022/12/11 (Y/M/D). Go to https://your_vps_address:8083/list/server/ and check which version you're running.

Obviously, there is something very wrong with my updates. I'm still at version 1.6.8; that must be a year or so that no updates have been done. Any idea why updating stopped at version 1.6.8? And, the important question: What should I do? There are Hestia scripts to update the whole manually, I think. But, how secure they are? I mean that, for me, it's really not the moment to risk that something goes wrong and that the website will be down... Thanks for any suggestion.

 

[Kairion]

26 minutes ago, allu62 said:

So, Uptime Robot works well? Maybe I'll try it... But can it really monitor all services? I mean all those listed on the Hestia page (including iptables, spamassassin). Do these services all run on a given port and how to do to find out what port it is (sorry, my admin knowledge is not lots more than zero)?

I don't think iptables and spamassassin could be checked with Uptime Robot since it just check if a desired server/port can be accessed externally and both are internal services (so either you have to check if they're running by yourself or your panel should do it for you), but I may be wrong.

30 minutes ago, allu62 said:

Obviously, there is something very wrong with my updates. I'm still at version 1.6.8; that must be a year or so that no updates have been done. Any idea why updating stopped at version 1.6.8? And, the important question: What should I do? There are Hestia scripts to update the whole manually, I think. But, how secure they are? I mean that, for me, it's really not the moment to risk that something goes wrong and that the website will be down... Thanks for any suggestion.

HestiaCP 1.6.8 has a known issue where updates and auto-updates with "v-update-sys-hestia" (default command for updating HestiaCP) are broken.

As a root user you should run:

apt update && apt upgrade

On your shell/terminal.

Another option can be found here: https://forum.hestiacp.com/t/update-hestia-v1-6-8/8608

About how secure it is: I can't really assure you it is 100% safe (always make a backup first) but it should work fine and without any issues.

 

If possible, I'd suggest you to migrate to another panel. Don't get me wrong, HestiaCP is a really good panel, but it does lack some features (e.g. IPv6 support) and since it doesn't have a commercial version to back its development everything is done by volunteers on their free time and thus would require some knowledge from you to fix things by yourself (like that update bug on version 1.6.8 would require you to keep checking HestiaCP website/forum/GitHub).

I did some tests and the most stable and easy to use panel I've found was KeyHelp. I'll make a post as soon as possible talking about some panel options to use on a VPS, but you can read a summary I made here:

 

[infantex]

On 1/21/2023 at 7:47 PM, Kairion said:

Could you share what settings you're using on Uptime Robot?

For the website:
   Monitor type: HTTP(s)
   Friendly name: infantex.com.mx
   URL (or IP): https://infantex.com.mx
   Monitor interval: 30 min
   Monitor timeout: 30 s

For the mail servers:

Monitor type	IP (or URL or Host)	Port	Monitor interval	Monitor timeout
Port	infantex.com.mx	465	10 min	30 s
Port	infantex.com.mx	995	10 min	30 s
Port	zaldivar.mx	POP3 (110)	30 min	30 s
Port	zaldivar.mx	SMTP (25)	30 min	30 s

 

[infantex]

On 1/21/2023 at 8:38 PM, wolstech said:

If you want to use the CF proxy feature, set Uptime Robot to watch the server's host name instead of your own domain (it'd work if he points it to tommy2.heliohost.org instead).

OK, I will change the settings to monitor tommy2.heliohost.org's ports 465 and 995.

But, can I be sure that if Tommy's mail is up so are my domain's?