Cloudron Forum

Apologies for any confusion. To clarify, I am suggesting that we set up an ingress proxy on our network with Cloudron operating behind it.

Similar to what happens with a Cloudflare Proxy install.

Hello to all,

I was testing a proxy setup (to reduce our needs for IPv4), and I noticed that for WildCard DNS, Cloudron would try to get from the nameserver his own IP.
But this will always fail because the DNS point to the proxy.

Would it be possible to validate with HTTP?

@nebulon i think that directus 10 should be not an autoupdate because of the license chage

@RazielKanos
The main question is scale; even 200 wp is a small group, especially if you don't let your customer install what they want from the store, and the store that you have don't have custom code in it, that needs to be tested, or are regionals store that doesn't need to scale.

And in any case, wp is not modern, it tries to stay up to date, but it is a monolithic software, which means you have disadvantages.

• Lower security because the frontend has access to the DB, which allows more exploitation.
• Front-End and backend are together; this doesn't allow you to scale quickly/easily.
• Plugin compatibility with different versions of PHP.
• WordPress is terrible in rollback, you need to use a backup.
• Working on multiple branches is not possible, in some cases not even staging is possible.

Maybe in some cases, the advantages (easy to find a compatible provider, plugin ecosystem, easy to find documentation, visual builder) make it a good option for you, but you can't say that is not a hard solution to maintain compared to a JAMstack solution.

But if you ask me what you should start with, if you want to build an e-shop, I would go on JAMstack almost in all cases, vue and svelte are super easy to learn and you can do a lot with almost 0 understanding of JS or TS, copy and pasting from the documentation.

@jdaviescoates
WordPress plugins are a mess, you need to update them weekly (for security), but that often introduce incompatibility or new bugs.
Mainly because plugins don't often document how they work and how they interact with the WP-Core.
We have noticed WooCommerce websites with a pretty standard install (wpcore, woocommerce, divi, wpml, some woocommerce premium plugins) getting completely unreliable after just 1 year of updates.
And due to lack of documentation, debugging is hell.

@timconsidine
any jamstack solution with Strapi/Directus + Paypal or stripe, and a front end that you can ask to build to a freelancer instead of relying on a hard-to-maintain WooCommerce.

@andreas
That front end code and the code server ware out of sync.
So any file change on the front end was not updated un till a commit was done.

For example if you have a Vue app, vite will not update your dev build untill you didn't commit

@andreas
Ohh they finally push it to prod, it was in testing with no file sync from February.

@JOduMonT

ECC now has mostly the same price as non-ECC, and AMD CPU support them for years.

But in general is better to have ECC if using FS that takes advantage of that, BTRFS, ZFS.
Or if you use TMPFS.

But if you care about latency then normal RAM is better, ECC slows down communication.

@girish said in Add Pre-Backup and Post-Backup Hooks:

Seems to not make sense.

That's normal for ERPs, almost all ERPs require full access to their database.
This is mostly due to the Modular idea for modern ERP, which implements different DB for each module like a microservice.
And communicate between with API or Functions Hooks

@marcusquinn
idk what tech is behind their stack, but if it is just HDD with no NVMe for caching metadata or any kind of index.
100% that it will be slow if they grow too much.

@1Psychedelic
do you have proxmox in Bridged, Routed, or Masquerading mode?
Because some time is needed for some custom config base on that before a server is able to access the WAN(aka the internet).

@BrutalBirdie
fantastic, so we don't need to do crazy remapping/system link to go around the read onlyfs.

@nj
that is correct if you have ZFS or BTRFS, that can do a snapshot instantaneity.
But cloudron uses a simple compression or rsync, which means that from when it starts and when it stops it takes time and in that time, data could be changed and the relationship could come out corrupted.

postgress in their documentation, for example, suggests that if you want to use an fs snapshot you should stop the DB.

@nj said in ERPNext - cost-effective ERP solution:

Cloudron totally takes care of backup and restore and I have already tested it

Not really, cloudron takes a snapshot of /app/data, and your database I guess don't create a dump before that happened.
That's not really stable as a solution.

If you could like get a dump every x time, that could be a solution, even if not perfect

@LoudLemur
The issue is that tech change so fast and maybe my information are old.
But ty

@nj
What you suggested is actually really hard to backup.
And it's not the best to have too many process in the same container.
Docker container differentiate from other container tecnologie to be single service container, even if they support multi services is not the best, especially for db.
Due to resources management especially.

@LoudLemur
There are many advantages that we don't speak a lot about, but there are some disadvantage too:

What you say in odoo is mostly due to the nature for a read only filesystem.
That don't allow to save data outside 4 directory, very comune for webapp.

Nextcloud is a webapp and is build as it, so it's more flexible to be restricted in directory location.

Odoo is an ERP and is actually a quite good one, and ERP are develop differently, they are not webapp.

Second problem is that Cloudron don't use Pods or Compose, but single container that mean that every app can be run in a single container.
And secondary been Cloudron base on docker and been implemented as the main controller for it, can cause problem if others panel interact with docker.

That's is the reason on the incompatibility for wings.

@robi
On that, I don't have the data or experience to talk.
But 100% sure you can't use wasabi latency, I think smb or better NFS should be a better option, but I'm not an expert.
and Using so much ram for it, im not big fun about it, but I would have to try it before to have a real feedback

@robi
yes, but then your image is used to run all your apps, which means that all those apps to run will have to get data from s3.

Remember that that image is not replicated anywhere else in the Filesystem, the container read their dependency from that file.
It would be like booting your server from s3, not a good idea, but that doesn't mean that storage footprint optimization is not possible in another way.