Site refusing to connect
-
Hello, my cloudron server refused to connect via the url which I normally access.
This is the error when I tried to log into the control panel
unexpectedly closed the connection.
Try:Checking the connection
Checking the proxy and the firewall
ERR_CONNECTION_CLOSEDI had not made any changes to it, and thought I would see if a reboot would assist.
It made no difference, so I stopped the server and then started it again to see if that would make a difference.It did not.
So, I contacted my developer who advised me to log in via ssh and perform some updates but I made a mess of things because I told him it was my runcloud server, not my cloudron server, I must have had a blonde moment being anxious not being able to access my primary invoicing server.
Subsequently I did an update when the information clearly told me not to, which I read after I had done the upgrade. I am normally not this flustered but I didn't know who to contact and it has 3 invoicing systems on there which I can no longer access.
To add insult to injury, I couldn't restart the server because it couldn't find the IP address - the internal one, although it looked like elastic IP had been configured - something is amiss.
So I have something I don't know how to fix.
It's an ec2 instance which I've had running since September 2022 without issue, in fact it's been flawless, but for the life of me this error has got me stumped.
I don't modify things, I had not done any major updates except updating the apps via the panel.
Is it possible to bring it back online?
I haven't got any idea on how to start the process.When I try to use the sudo reboot option this is what I get -
sudo reboot
sudo: unable to resolve host ip-xxx-xx-x-xx: Name or service not knownHowever, the IP is connected to the account and is an elastic IP when I look at the aws settings.
I truly do not know why it stopped. I was using it one day, and then went to access the site the next day and I can't find any posts regarding this type of error, support@runcloud.io only does billing.
So I'm wondering - is there any possibility to extract backups from this instance and do a brand new one?
I would be very grateful for someone with wise eyes and mind with cloudron to point me in the right direction because I can't even articulate what I'm meant to ask for to get answers I need.
Thank you
-
-
G girish marked this topic as a question
-
This is not a stupid question at all - it's stupid person - ME, because I have BOTH runcloud and cloudron in my AWS setup. What I did was email runcloud, what I did was relay cloudron to my developer, who instructed me remotely on what to do as he was unavailable.
And yes, another rookie mistake - those apparent blank things are nothing but me putting two > > next to my comment. Gosh it's been a long day and I am sorry I didn't make myself clearer.
-
Servers do look very similar on SSH, so such things can happen
Given the likely unknown state of the system, I would suggest to simply restore the whole from a backup https://docs.cloudron.io/backups/#restore-cloudron this would likely be safest to avoid unexpected side-effects.
-
@nebulon Ok so I obviously have made another rookie mistake - backing up only the apps on the server, and I have an instance from 30/8/2024 when I expanded the hdd, but none apart from that.
I used to download some of the instances, but this was setup for me, and I thought I was doing the right thing backing up the apps daily, and little did I know that it was not correct, given that I also backup pretty much everything.
Is there a way I can use ssh to get to the folders where the backups are held on the drive to download the latest backup files, spin up another instance and then reinstall from that?
Thanks for replying. -
@nebulon Thank you - I will research how to get to that folder because all I can see in ssh is the amazon one and the lxd folder. Then there are a couple of other folders but I can't work out how to navigate to that folder. Yeah I'm not an ssh expert because although I'm familiar with working with a dos interface, there are obviously some subtle differences in the navigation of a linux / ssh instance.
-
@nebulon I used Chatgpt to find out how to safely ssh into my cloudron server and download what I thought could have been a backup file but it must have been the master file backup when it was installed in 2022.
I then asked chatgpt to help me search the server for any tar.gz files after December 1, and there was only one backup file there which I thought was confusing, and then I remembered, I had them all setup on cloudflare to backup and I FOUND MY BACKUP FILES.
I'm not sure how to approach this now - should I spin up a new cloudron snapshot and install from backups? I had issues in the past whereby I couldn't install snapshots because they were different versions to the installs to restore over, so I guess my next question would be -
IS THERE ANY WAY TO RESTOR A SNAPSHOT without having to install the app - eg, invoice ninja and then tryh to restore an older version? I know this sounds convuluted - but is there a way to simply restore a snapshot minus the initial install?
I have learnt so much today. .. One being - Yes I have backups, and yes although it's inconvenient, I won't have lost data. I don't know whether to laugh or cry.
Thanks
-
@berryd If I understand this right, you are using Runcloud on a AWS whatever in order to run Cloudron... right? If so, you've misunderstood deeply. Cloudron REPLACES Runcloud. Cloudron IS essentially Runcloud. Whatever Runcloud says it does, Clourdon ALREADY DOES IT without the need for anything else. Are you really running Cloudron as an app within Runcloud?? This is the main reason then for your confusion, as the two are essentially the same thing. Both manage servers and run web apps. CLoudron is not just a webapp, and trying to install it on a server that already has something, almost anything else, will cause problems. Remove RunCloud, and cancel that subscriptoin. Then go back to https://www.cloudron.io/get.html and https://docs.cloudron.io/installation/ and reinstall ONLY Cloudron on a fresh server. Then, enjoy!
-
@scooke I use cloudron and runcloud - I have lightspeed and ec2 instances. I made the mistake of when I discussed with my developer I used the word 'runcloud' instaead of 'cloudron'.
I run both - I am running cloudron on an ubuntu ec2 build. This is the one with the issue.
I have located my backups on cloudflare.
I just keep having a brainfart between the words 'runcloud' and 'cloudron'.The reason I ended up using cloudron was because of issues with laravelle and invoiceninja where due to how runcloud works, it was not possible to get working, so my dev walked me through spinning up an ec2 instance and connecting to cloudron and using the app to build it. It's absolutely fantastic.
I don't know why I keep on getting it mixed up... It must be my left and my other left, but I definitely know what I have built.
Thanks for replying.
-
I solved the problem -
Purchased another hosting plan from a provider who supports runcloud
Creating a new instance
Imported in all the data from the backups
Redirected the IP addresses after connecting to cloudflare
Works like a charm
Used ChatGPt to walk me through some of the ssh setup
... I learnt one thing... If you are ever doing this, leave things alone for at least 20 minutes before changing something. It was like a perpetual loop of not working and when I left it, came back an hour later, put the same strings in to connect to the cloudflare r2 buckets, it worked like a charm.I will be honest, I am quite happy I learnt a few more things. A fresh instance of Cloudron was the answer, I don't know why it happened in the first place, given nothing had been changed and one day it was broken after running without issue since Sept 2022, but now I have confidence in fixing things should it occur again.
Backups were really important. Thank you
-
J joseph has marked this topic as solved
Also, your screenshots are blank, so I can't make out which product you use.
