Cloudron Forum

@girish said in What's coming in 9.0 (was 8.0):

Multiple Backup Destinations

This is dope, can't wait to integrate this in our backup process !

Ollama should definitely be part of the app package, even if it's optional! We're extremely interested in this possibility and have a very powerful infrastructure on which Cloudron is installed. For companies like us, this would be a total game changer!
What's more, we think it's very important to always leave the choice open in this kind of project. That's one of the reasons why we chose cloudron.
What's more, we would use Ollama a lot.
Have a great day and thank you for this new feature.

Dear Cloudron Team,

I hope this message finds you well. I am writing to inquire about the current status and future plans regarding the multi-hosting feature for Cloudron.

We have been using Cloudron for both personal and professional purposes for over a year and a half, managing a wide range of applications. Our experience has been largely positive, and Cloudron has proven to be an invaluable tool for deploying and managing our digital infrastructure.

However, as our usage continues to grow, both in terms of user numbers and data volume, we are increasingly finding the need for horizontal scalability. The ability to deploy Cloudron across multiple servers, ensuring synchronization and seamless operation across these instances, is becoming critical for us. This feature would not only help us manage the growing load but also provide a robust solution for future expansion.

In 2021, there were discussions on the forum indicating that multi-hosting was being developed and expected to be released with version 7. Since then, we have not seen any updates or releases addressing this capability. Given its importance to users like us, we are keen to know if multi-hosting is still on the roadmap and, if so, when we might expect it to be available.

The ability to scale horizontally is essential for us to consider Cloudron as a long-term, viable solution. It would significantly enhance our ability to manage a large number of users and substantial data volumes without compromising performance or reliability. We believe this feature is crucial not just for our use case, but for many others in the community facing similar challenges.

Could you please provide an update on the development of the multi-hosting feature? Is it still being planned, and do you have an estimated timeline for its release? Your insights and any additional information on this matter would be greatly appreciated.

Thank you for your continued hard work and support !

Hi Lukas, I'm not sure what I'm going to tell you, but what I did on my own infrastructure, we did it completely differently, that is to say that on the first disk, the server one, we installed Cloudron, but on another storage volume that is not even a disk on the machine, which is an NAS that we mounted as volume on the machine and that we mounted as volume on Cloudron to put heavy applications. And it works without any problem on our side.

I think that as long as the volume is well mounted on Cloudron as a new volume, it will work without any problem. Or if it is detected natively by Cloudron as a second basic storage system, in this case there is no need to mount it as a volume I think. But it remains to be checked with Girish.

I use a lot of OpenWebUI, at least I try to use it even if it takes me 3 minutes to generate an answer every time. But it's a problem of infrastructure and not of application.

The problem is that everytime I set my advanced settings When I reconnect on the same machine or on another machine, the advanced settings are systematically reset And it's a problem because I like to have more than 10,000 tokens coming out everytime And keep the Keep Alive over 8 minutes, which is mandatory with the current configuration. And the fact of having to systematically go back to the settings before use to be able to reactivate these two parameters is quite annoying.

I would have liked to know if in the next patch it was possible to make sure that the advanced settings are not reset. Thank you.

Also, I take this opportunity to talk about this, it's not a very serious problem, but it's true that on the first version before the update there was no problem so it was nice. It's that the profile picture of the model no longer loads, I don't know why, but instead I have this one showing up :

Also, I take this opportunity to talk about this, it's not a very serious problem, but it's true that on the first version before the update there was no problem so it was nice. It's that the profile picture of the model no longer loads, I don't know why, but instead I have this one showing up.

Dear @crazybrad ,

Thank you for sharing your thoughts and experiences. We completely resonate with everything you've mentioned. Just like you, we've found ourselves disappointed by larger, more established service providers in the past. Cloudron's incredible team and supportive community have indeed been a refreshing change, offering us a more personalized and efficient approach to managing our digital infrastructure.

We strongly agree that achieving our goals with Cloudron will require a collaborative effort. We are absolutely prepared to invest in our future with Cloudron, whether that means contributing constructive feedback, sharing innovative ideas, investing our time in the community, or even providing financial support.

Cloudron's team has demonstrated an exceptional level of availability and dedication to their users, something we've not experienced elsewhere. This is precisely why we initiated this discussion. We genuinely believe in Cloudron as both a short-term and long-term solution and want to be proactive in ensuring its continued growth and success.

We're committed to helping Cloudron reach its full potential because we want to continue relying on it for our evolving needs. We envision a future where Cloudron can scale horizontally to meet growing demands, and we are eager to support the team in any way possible to make that happen.

My bad. Thank you @jdaviescoates

I searched and saw it but due to the creation date that was more than 5 years ago I didn't know it was still considered as active.

@crazybrad to address your points:

We don't use WordPress on Cloudron. Our WordPress sites are hosted on dedicated servers at OVH. I mentioned WordPress earlier to align with Girish's example, but it's not part of our Cloudron setup. We don't use Cloudron for a large number of websites.

Currently, we support around 40 users who utilize about 30 different applications deployed on our Cloudron infrastructure. Many of these applications are heavily used, although they don’t consume an extraordinary amount of resources individually. However, we aim to maximize our self-hosting capabilities with Cloudron, integrating many applications into our workflow, which results in significant usage.

We plan to expand our user base and the load on each application in the coming years. For example, applications like OpenWebUI consume substantial resources when in use. Even with a single user, the resource demand is high, and this is compounded when multiple users are active simultaneously. OpenWebUI, specifically, would ideally run on a separate infrastructure designed for AI, yet remain accessible to our users without requiring new accounts or different admin panels.

Switching panels or remembering different subdomains for admin tasks complicates matters since we’ve already trained our staff on using Cloudron and each application within it. Simplifying administration and user experience is crucial for us. Just like the possibility of not being limited by the constraints of the monoserver-only self-hosting, so we can consider Cloudron as a Long-Term solution.

Additionally, when you mention having "given up self-hosting as not being worth the risk of a catastrophic failure," are you referring specifically to using WordPress with Cloudron, or does this include self-hosting in general?

Hello,

I have an infrastructure currently with two disks, a classic 500 gigabit disk and a 500 gigabit disk. I have already done some research and it seems to me that I have not found my answers to the questions that have already been asked on the forum. That's why I'm allowing myself to create a topic.

I have some users who use my cloudron with me and this generates a lot of data. First of all, I have to know that I already have my docker that consumes more than 50gb. Regarding this subject, I saw that there had already been a topic that had been created on this subject and I do not seem to have found any solution or things to do about it. My only question regarding the docker that consumes more than 50gb is: will it continue to increase indefinitely or will it regulate and delete files as it goes? And that after a while it will stop increasing? That was a small bonus question.

My real question is, knowing that I have a first 500GB storage disk on which my Cloudron is mounted, and a second 500GB disk which is currently set to RAID, I wanted to know if I would set my second disk as classic storage, would Cloudron be able to use this second disk as a normal disk, that it would be taken into account in my backups, that my applications would not be a problem, etc.
Is this manipulation feasible, compatible with the installation of the Cloudron? Will it pose a problem at some point? If I have data from an application on the first disk then the first disk becomes full and it goes on the second disk, will the applications have no problem going to look in disk 1 or disk 2 depending on the requests? I don't know at all how it works, that's why before doing the manipulation I prefer to ask if it is advisable and feasible or do I have to change infrastructure and go on an infrastructure that has two bases, much more storage to solve this problem.

Same here. On all my stations. Plus the fact that the WebDJ doesn't work.

@Aizat Same here. And we could be very interested in having this app in the Cloudron App Store too.

Hello @girish ,

Thank you for your detailed response. Yes, we are indeed talking about the same thing regarding horizontal scaling or multi-hosting in the Cloudron context. The ability to have a single dashboard for managing apps deployed across multiple servers is exactly what we are looking for.

We are pleased to hear about the imminent release of Cloudron 8 and the plans for Cloudron 9, especially the focus on improved backup strategies. These updates are definitely welcome.

The solution you proposed, where we can manage applications on different servers from a single dashboard, sounds promising and could indeed be a viable long-term solution for us. This approach allows us to distribute the load manually across multiple servers while maintaining centralized management, which addresses many of our scalability concerns.

Currently, our primary Cloudron instance manages a substantial number of applications and user data. As you rightly mentioned, having multiple instances requires remembering different subdomain dashboards, which complicates user and admin management. Moving towards a solution where we can deploy applications like WordPress on one server and NodeBB on another, all while managing them from a single interface, would greatly simplify our operations.

I will review the documentation you shared on using an external user directory and look into the upcoming changes to OIDC for single sign-on across instances. While this current workaround helps, having a more integrated and seamless multi-hosting capability would be highly beneficial. This could be a short-term solution, until the real feature we've just been talking about is implemented. But we don't think it's an ideal long-term solution for us.

Having the ability to manage multiple servers from a single dashboard, as you've outlined, would provide us with the scalability and flexibility we need. We are eager to see this feature implemented and would appreciate any updates on its development timeline.

Thank you again for your continued support and for considering our feedback. We look forward to the upcoming releases and future developments.

Every time I connect to SSH on my machine, I realize that I always have these updates that are there, these 19 updates that could be applied immediately. When I run apt list --upgradeable, what happens is that I see that I have 9 urgent security updates. I know that I am not supposed to update by apt, but I wanted to know what it was about these security updates.
Am I really supposed to not update theses ?

Because I saw that Cloudron was supposed to do the security updates automatically, but when I updated my Cloudron for the last time, it obviously didn't do it.

I tried to look for the answer in the forum but I found nothing on the updates of the OS of the machine that supported Cloudron. I only found things regarding security updates that concerned applications but never on the OS of the machine. That's why I'm asking myself the question here.

+1 for Appflowy. Very very high value added.

@crazybrad I understand your point about WordPress and the necessity of specialized tools for effective plugin management, backup, and update procedures. Your approach of using third-party hosting for WordPress and your tested backup solution makes a lot of sense, especially given the critical nature of ensuring reliable backups.

Regarding OpenWebUI, it's indeed a resource-intensive application, which is why we’re looking for ways to distribute the load more effectively while maintaining centralized management. Your insights on best practices for WordPress, especially the backup verification process, are very interesting. We’re exploring similar approaches to ensure our backups are reliable and that we can restore quickly if needed.

I have a server that has a lot of RAM and a good processor but very little storage. I have more than 40 applications running on this server, and it turns out that I have a lot of data and I transferred all the data from my applications that have the most gb occupied on a new volume which is a NAS-HA server from OVH and the applications work perfectly, all the data are in the right places, it works great, it's perfect.

My problem is that now I can't say if it's systematic or if it's only on some backups but this message comes back very often telling me "backup failed" with this error. First of all, during my first research on the forum, I realized that someone had said that in the documentation it was said that the other volumes are not backups. This is a problem for me, if there is a solution for backups, it would be interesting for me that it is included in the server backups because there is a lot of data on the new volume storage. And that's where most of the data from my applications will now be stored. So if they are not backups, it's a problem. But when I went to see the documentation, I didn't find any backup mentions in the volume part.

But I would like to know how to solve this problem that is coming more and more since I added this new volume.

@micmc I tried, many times actually. Never managed to make it work by myself. Maybe I'm stupid.. This is why I use cloudron..

I don't know if it's a real bug or not, in doubt I prefer to talk about it here. I downloaded three different models and this error occurs on all models. I have 100GB of RAM allocated to OpenWebUI And my infrastructure has 12 cores. The fact is that even when I restart the application, etc. Sometimes it works, but very often the IA loads in the void leaving me on this page indefinitely :

I don't know if it's normal or not but OpenWebUI doesn't use any RAM at all, it only uses my CPU. If I believe my htop, the application uses less than 1GB of RAM when it is used.

Another thing, always if I believe my htop, even if I leave the application loading indefinitely and therefore consequently after a while I close either the chat or the tab, but cores continues to be used a lot.

At some point, I wanted to try the Llama2 70B model and I let it run for about 40 or even 50 minutes. At some point, I closed the page or refreshed it, I don't remember. And the answer appeared like magic. The model had not been written in my eyes, but it was as if the interface had not been updated and had continued to show me the load. But the answer had already been written and when I refreshed it, it appeared.

The problem is that the Llama2 70B was probably too big for my infrastructure, but the 7B's... I don't understand why they put so much time to charge. Knowing that sometimes they charge almost instantly.

Hello,

I'm coming back to you because this problem is still unresolved. I haven't been able to find a solution to get the backups working again.

And I can now confirm that it is indeed all the backups that are not working. Every day the backup fails.

Hello, I have been a Cloudron user for almost a year now. I am very satisfied with the simplicity of the solution. However, there is an application in particular with which I have encountered problems since the first time I tried to install it. It's been more than six months now. It's pretty much Language Tool. Indeed, I tried several possible configurations to make this application work on my server. The fact is that it never works. When I try to see if the text is correct, I always have the same error, saying "Oops, sorry, a problem occurred when checking your text. Restart the browser to solve this problem (405)".

I read the installation documentation and consulted the forum to see if there were people who had this same problem. I saw that there was a person who had a problem 405. The fact is that his solution does not work for me, knowing that I do not even have a secret. I tried with a secret, I tried without a secret. The page of my language tool is accessible from the browser without any problem. However, it seems that the extension fails to work with it. I have here the error screen that I have in French, I didn't manage to see it in English for you.

In second screen, I joined the page that is visible from my "language tool" link which seems to work, to be normal.

I would have liked to know if I could have help on this error As I would have really liked to be able to use this app, it would have been extremely useful for myself and for my employees. And as a short idea of the solution to this problem, I call on you @girish .

Note that at the beginning I didn't install any n-grams, But seeing that it didn't work, I tried to install two, the EN and the FR I thought it would work, I restarted the app but the error persists, I have exactly the same error So today I have two ngrams installed, the FR and the EN.

Thank you for taking the time to read my text and I hope you will help me find the right solution to make this app work.