Cloudron Forum

Switched to tarball, and it made a huge difference!

I use Netcup as my host, in one of their datacenters in Germany, and do my backups to Wasabi, also in Germany, using rsync. Even though my Cloudron installation is only 40 GB, backups take at least 6 hours. Six apps installed: Mastodon, Ghost, Adguard, FreshRSS, Uptime Kuma, and Umami. Guess the reason is a bunch of small files.

I went with rsync for incremental backups at less storage space. But I am starting to reconsider that, and perhaps switch to tarball.

What backup strategies do everyone else have?

Ah, that make sense. So then I better select Dry run to not have any internal checks fail, change DNS manually, and after that are good to go? The Sync DNS wont work either for Wildcard DNS?

Hi!

I'm in the process of migrating from one VPS to another, and while waiting for the backup of my old installation finish I read through the docs once more and realized that there is one thing I don't understand: What does "Dry run" do?

I get that this is a feature that let me test my Cloudron on the new server before pointing the DNS to it. What I wonder is more like how Cloudron can make those changes in the first place?

I'm using the Wildcard configuration, and in the settings at my registrar point the A records to the IP address of my current VPS. How could Cloudron change those settings if I don't check the Dry run option?

What am I missing?

Hi!

The media cache of my Mastodon installation is growing to big for my VPS. I'm considering moving the media storage to S3 instead of upgrading to a larger VPS. But I can't find any instructions on how to migrate the existing local files (that is the ones me and the other users on my instance has uploaded) to S3. Anyone who has made this and can share how?

Netcup's DNS servers are configured for my Cloudron instance.

I'm on 8.0.6 since Tuesday morning, which incidentally also is when those problem started.

Should I use resolvectl from Uptime Kuma's web terminal?

@joseph Yes, this is what I get using the web terminal.

But could it be that the DNS is going down intermittently?

I've used Uptime Kuna without any problem since April. But for the last 24+ hours I suddenly have started to receive tons of these:

The affected hosts are on .se, and one .social. Some of them are hosted at the same box I run Cloudron on, some on others.

I see that it's not the same error message as in the OP, but similar enough to be related?

According to my logs, my Cloudron was updated to 8.0.4 on August 28, and 8.0.6 yesterday morning. I don't know for sure, but from the graphs it looks like the problem started as soon as that update had been installed. This is what the week graph looks like for or the monitored services with problems:

I rebooted the whole system, and it seems to have solved the problem.

I posted about a problem with Castopod earlier today, but I just realised that my Cloudron instance might have generic cronjob problem: For a couple of weeks, my FreshRSS installation has stopped updating automatically. I've to make a manual refresh every time I want to check my feeds.

Could it be that my Cloudron have some app-wide cron problem?

Thanks!

Hi!

I'm trying to import a podcast into a fresh Castopod installation. I've pasted the URL to the RSS feed, but the process is stucked on 'queued'. Searching Castopod's Discord, the proposed solution is to make sure the cronjobs are properly configured. But as far as I know, Cloudron usually adds all needed cronjobs when you install a new app.

Any ideas for what to do?

Thanks!

Ah, thanks! I hadn't seen the app-mail-folders before.

Got this reply from Storj @girish @nebulon

Could you change some settings in this software to use a delimiter for prefixes? By default it's `/` (a forward slash). Without delimiting you will hit the listing limit if you would use S3.  
As an alternative you may use a native Storj protocol, it doesn't have this limit, but I'm not sure that you can use it in this software.  
What's supposed to do? if backups, then I can suggest to use a normal ones, who can use prefixes correctly (and may reduce your storage costs by using a larger segments (less or equal to 64MiB)): [https://docs.storj.io/dcs/third-party-tools#backups](https://docs.storj.io/dcs/third-party-tools#backups)  
   
Our S3 compatibility you can see there: [https://docs.storj.io/dcs/api/s3/s3-compatibility](https://docs.storj.io/dcs/api/s3/s3-compatibility)  
However, there is a limitation, how much objects you can list using the not limited prefix (like a prefix without a trailing `/`).

While researching some other backup problems, I just read the support docs and have a question about restoring email from backups: How would you do that when you have encryption turned on?

@nebulon Sure – if I only can figure out how to see how many objects there are. Do you know?

@girish Yeah, using rsync. Can try tgz once I've checked in with Storj's support.

As of this morning, my backups to Storj suddenly fails. This is the error message I get.

Error listing objects in snapshot/app_0c3fb7c9-77c7-4d96-a90a-b75b5d1562bf. Message: ListObjects(V2): listing too many items for gateway-side filtering using arbitrary delimiter/prefix HTTP Code: NotImplemented

Any ideas?

Solved! With a push monitor in Uptime Kuma and a curl script on the computers.

Hi!

I'm using Arq Backup for a couple of clients. It has an option to send an email when a backup job finishes, which is good. But I would also like to get a notification if a backup job hasn't been executed for the last X days.

Is there any app on Cloudron that can be used for this? Assign an email address, and notify me if nothing has been received in a defined timeframe?

Edit: I just realised that Arq also can execute a script after a backup, so anything that could alert based on missing pings using GET/POST etc would also work.

So the best option for now is just letting them knock on the door? And perhaps check the logs every now and then and add IPs from there to the blacklist?