[Solved] Hestia CP inaccessible on my VPS

[Krydos]

3 hours ago, allu62 said:

Remains a problem with hestia-nginx: in the updates section, version 1.23.0 is marked with a yellow exclamation mark. Does that mean that update failed?

I ran the updates for your VPS. There were 150 packages that needed to be updated.

3 hours ago, allu62 said:

What is the best action I can take to resolve this?

I logged in via SSH and ran the commands

sudo apt update
sudo apt upgrade

Which you can do yourself occasionally in the future if you want to. If the update asks you to keep your existing config or overwrite it I would select N for keep your existing config.

[allu62]

Sorry that you had all this work because of me. I should take the administration of my VPS more seriously and regularly checking if updates have been done, checking the logs and graphs, etc. Thanks a lot!

Two questions (curious and trying to understand):

1. Have you any idea how such problems can occur? If there weren't any updates done, all remained as it was and as it worked how can it suddenly stop to work? Perhaps, because some components did do an update? Normally this would mean that all dependencies would also be updated. Or is the Linux "perfect package system" not really perfect and the system can break due to packaage incompatibilities?

2. How did you do to connect to my VPS? Using a sudo account that you created when installing Ubuntu?

Anyway, I'm really grateful. Thanks to all those who, during the last years, helped me to keep my website running!

 

[Krydos]

I haven't used Hestia enough to know if it does automatic updates or not, but I'm guessing even if it does update automatically it only updates its own components, not the rest of the operating system. Running apt update && apt upgrade should be done yourself maybe once a month at the least. Be ready to fix something if it breaks though so don't run the update unless you've got an hour or two to fix anything that goes wrong.

When the VPS is created the VPS admins, currently only Yashrs and myself, have root access to the system. You can remove access if you want, but if you ask for support we can get root access again easily enough. Leaving the access we currently have allows us to save about 15 minutes though per support request. As we get more VPS subscriptions we may train another admin to help, but for now it's just the two of us. Technically all of the root admins, Ashoat, Byron, Wolstech, and myself, have access, but since they're not all trained it's unlikely they'll ever need to. Our policy is we don't log in to VPS unless we have a good reason to believe that they are breaking our terms of service, their VPS has been hacked, or if someone asks for help.

[allu62]

So, if I understand, Hestia may break the system by doing its updates? I think that the actual web interface has an option to disable automatic updates. Wouldn't that be the better choice? After all, at least I suppose so, doing a system update would also update Hestia? But, I guess this is only the case if Hestia is an official Ubuntu package? So, perhaps the best practice would be to run apt-get and then run hestia-update manually?

Concerning your root access (does that mean that Ubuntu has a user "root"?) to my VPS, that's more than ok. First, there is no reason not to trust you; second there is nothing that's not anyway publically accessible, and third it may be more than useful for me (and letting you lose a little bit less time).

 

[Krydos]

The most common issue I've seen with Hestia autoupdates is when the old Hestia process refuses to exit for the new process to start. This is the type of error you will see when that happens.  

On 4/13/2023 at 4:01 AM, allu62 said:

Apr 10 12:25:23 vps43.heliohost.us hestia[187710]: nginx: [emerg] bind() to 0.0.0.0:8083 failed (98: Address already in use)

The old process is already using the 8083 port so the new process can't bind to it.

11 hours ago, allu62 said:

does that mean that Ubuntu has a user "root"?

Yes, all Linux systems have a user named root.