Question Can Plesk do it?

[orion]

I have used Plesk many, many years back, I think it was still before SW Soft purchased them and it was decent, not better than cPanel but good. Then with SW-Soft, when they had the brilliant idea to advertise Virtuozzo on the login page and header (plus all the things you could not change for branding and the constant pushing of services to my users) I decided to move to cPanel like many others back then. Then came Parallels and it stalled. Then Odin...

I never really required the cPanel support extensively, but I have to admit they are not good, but amazing. One of the best support's I ever had from any software company. I don't expect Plesk to compete there and to be honest, I don't even care about that. I don't expect someone to offer 24/7 support for that price, but my questions are more related to where Plesk is moving now. And instead of offering lousy support, you should better work with partners and companies that do that instead instead of hurting your brand with outsourcing to low techies that don't know basic Linux migration procedures. (yes, I read horror stories from your support.)

It seems now you are finally free from Odin now but for how long? Really. How many times did Plesk changed hands so far? I do like the new company and brand. It seems you are only focusing on doing one thing and doing it right vs too many things poorly. Plesk Onyx seems like a long wanted refresh, and I like what I see, but I also see some of the old Plesk trends around that don't want to die...

I downloaded a trial and while it seems you are a bit more open to branding now, it still looks very boxed and limited to cPanel and other softwares when it comes to changes. Can I now finally make my own theme entirely different to Plesk? (different colors, icons, text, everything???).

Plesk is 100% encoded so changes in the past where a paint in the butt or impossible in most cases. Plus they tried to push their marketing and branding to your end customers. This is something I find unacceptable on a paid licensed software towards service companies. In particular, when open source is so strong today, there is no reason to be paying to have something that you cannot change and where the company is advertising to your end users that you are trying to keep. So said this. I can change Plesk to the point it does not look like Plesk regarding looks? How about disabling the store or extensions or things I don't want pushed down towards our users?

This is something that I also didn't like about Plesk. They really try to make a quick buck with extensions and every single extra. Makes your software feel poor and the popular opinion is that the barebones is useless and requires alot of extra paid stuff. You are the only company still charging per language file. At least charge for all languages together or make a decent price. Once you add 2 or 3 extensions, Plesk is by far more expensive than cPanel, 3 times more. And no, I don't want or need the power packs. I don't care about your antivirus plugins or Cloudflare or anything like that. There is no reason why I should buy a pack in order to get cgroups or remote Docker.

 

[orion]

Speaking of this. While it's very nice that you guys added support, this seems like poorly done. In your documentation, you even advise that Docker should not be used for customers as container security is inadequate (and rightly so).

So why are you then letting users host them locally on the same server? What exactly do you think the purpose of this is? For an admin to test something for a few minutes? If someone wants to launch a Docker container it's to provide something to someone (container access or service access that runs on them) and you can't give Docker access to any service or customers unless you put those customers containers on its own enviroment (VM, VPS, etc.).

So basically the only option is the remote plugin (again extra charge). Except for the plugin only supports 1 remote server? Really? So you can connect 1 single VPS, to offer Dockers to 1 single customer? Seems like poorly implemented. So a customer that wants Dockets, now costs you $5 a month, and you can only have one. Do you really expect your customers to compete with Amazon and Google like this when can get a $3 VM with them? Again, poorly implemented. It seems like Plesk just wanted to say, we have Docker first. Look what Rancher is doing if you want ideas...The only option is to give Docker remote for free, and please, allow more than one server connected if you want it to make useful...A feature like this is just marketing but not really usable.

Same is true with DNS. It was shocking that Plesk didn't had any type of DNS clustering. Seriously, what kind of companies do you expect to be providing different DNS servers for every single server? Ok, you fixed it now with the slave plugin, centralized DNS clustering seems to be here, but is it stable now? It seems like a nightmare after a bit of reading about the bugs and issues it has. But at least, great, you finally realize what a service provider wants.
How about changing the root or /home path for a particular subscription or customer? Is this now possible? What if I want some accounts to be stored in a different storage in a server? Can this be done now?

Multi-server, excellent, but not affordable. Really? $100 per server a year? Seems like a big piece of what the whole Plesk software costs per year. At that pricing, you are better just using something like Ansible or another automation tool. Again, someone at Plesk just makes pricing up without actually researching what tools companies are using what the competition is doing. Anyone with a few servers would be dumb to be paying that when there are better tools that do far more and better. Not even mentioning that you can do most of that with scripts and automation. I would not be paying even $10 a year based on that extension does right now. In particular when it just increases your costs again, when the industry is lowering per customer pricing. It seems everything you add to Plesk just increases it. This is the big and number one reason people stick and move to cPanel. One pricing, everything included. Not gimmicks.

Don't get me wrong. I WANT to migrate to Plesk. But while some things look really great now, some things still tell me not to do it. I have the impression I will be locked once Plesk decides to raise prices or just kill features because they think they are way too cool in terms of features. Speaking of that again. In the past, you offered Presence Builder with the highest edition. That product is EOL and your Plesk hosting edition costs the same. How come? Do you plan to add something unique to that edition or replace it with another sitebuilder solution?

 

[orion]

One of the reasons some people moved to Plesk, was actually that software. No joke. I used that software with cPanel servers as well, and it was by far the most decent sitebuilder available (not great, but decent from all the garbage out). As companies lose customers over to Weebly and Wix, not having something to offer users that have no idea about WordPress and just want a simple drag and drop interface was a strong selling with Plesk.

I know some moved to Plesk just to be able to offer that. Now, this is gone. Are you planning to build something in the future similar vs newbie users that require a stupid simple sitebuilder? You can't believe how many companies I know that are using something like that. And since its plain HTML there is nothing to maintain or upgrade either, you can't hack HTML !!! And with the new rise of static CMS (no datbase), you should really look into having a sitebuilder for static sites.

I'm sure this post will help a lot of other persons and some official reply would be welcome. If Plesk does things right, they can actually go big, and I mean colossal, they can eat into cloud vendors if they offer the right tools to their customers. They still have to remember that their clients are companies and mostly service providers, so they have again to provide services to end users. The last thing a vendor wants is just one more of the same. This is why its so important to have the right pricing (you need to take into account all costs, with extensions) and to be different regarding products, services, and branding. What if I need to integrate my own ticket system into the panel, or my own monitoring stats? Sure, everyone has an API and extensions now.

What you need to understand is that shared hosting as we know it is dead. The only reason its still working is because CloudLinux is solving some of the long problems, like security and resource charging. Except that again is another product. You can't seriously offer script access like PHP or Ruby or a shell account to anyone without proper isolation. None of the panels do that, and while Plesk now has cgroups (a good step towards the future), I don't see them replacing CloudLinux either, which means Plesk as barebone alone again is useless for any serious company. (if you know how shared services work you will understand what I'm talking about regarding security).

So unless Plesk plans to dockerize customers or start using namespaces and file system virtualization to cage users into their chroot environment's, you have no future because hosting shared companies have no future. The newbie webite market is taken away by companies like Weebly, no solution with Pesk for that. And the bigger customers need powerful resources and security for which shared hosting is not a solution either.

I hope someone at Plesk that actually knows and understand how hosting works replies here. Not about what Plesk does today. We want to see a roadmap or your idea and vision to the future. That is what a company wants to use your software which you should consider a platform and not just a software. People want HAproxy load balancing, fault tolerance, high availability, none can be done with something like Plesk or cPanel without changing it to the point you don't even need the software anymore at that point. The only thing I see you have done in that aspect is allowing remote MySQL databases (which everyone else has as well), that is hardly enough for a more sophisticated environment.

Plesk community. I want your real honest opinion. Not trashing cPanel or how great Plesk is. I want to see where you think this is going towards the future and what needs to be improved or is planned ahead. If your customers fail, you fail. And everyone here should realize that in order to compete with something like Google, Amazon or Microsoft (all 3 offering cloud services) you need something bigger, stronger and better. Providing rock solid email services, a great sitebuilder, a platform for easy and click launch applications, you get the idea. Whatever you are thinking Plesk team, think bigger.

 

[Sergey L]

Hi, and thank you for your feedback.
There were many topics brought into discussion, we will try to handle one at a time.

I am not sure what sort of branding customizations you lack comparing to cPanel. If you can show a particular need that you were unable to fulfill, then we can talk about the particular problem and perhaps plan an improvement

What I can see relevant to this problem now:
- you can disable a lot in Plesk UI through panel.ini file and if you don't like Extensions particularly you can disable extensions as a whole or individually
- you can also build your own view of Plesk on CSS and images level. Yes, you might need to adjust it with new versions of Plesk coming
- you can build your own texts too if you build your own language files. As this is requested only once every couple years, we don't have a public documentation probably - yet it is not really difficult for anyone who knows Plesk a little bit. A piece of advise - if you develop it as an extra language (rather than replace English locale), it will be tolerant to future updates as texts missing in your locale will be taken from en-US language.

You can checkout documentation for other options and let us know if you don't find something essential

 

[Sergey L]

Regarding being encoded (rather than open-sourced) and changes to perform in Plesk. The Extensions (which you didn't like?) are designed exactly as a response to the need of heavy customizations in Plesk. If you check out our SDK, you might find out that you can extend or modify Plesk code to a large degree, especially if Javascript is brought into equation. I can name some well known hosters who have used extensions and aforementioned custom themes to modify Plesk look, feel and even behavior to the degree that the one might be unable to recognize Plesk. A lot of extensions are already available at github opensourced, so you can use their code as example.

I hope that helps, and let us know if you would have any particular problem or lacking need in SDK.

 

[Sergey L]

This is something that I also didn't like about Plesk. They really try to make a quick buck with extensions and every single extra. Makes your software feel poor and the popular opinion is that the barebones is useless and requires alot of extra paid stuff. You are the only company still charging per language file. At least charge for all languages together or make a decent price. Once you add 2 or 3 extensions, Plesk is by far more expensive than cPanel, 3 times more. And no, I don't want or need the power packs. I don't care about your antivirus plugins or Cloudflare or anything like that. There is no reason why I should buy a pack in order to get cgroups or remote Docker.

I am not sure if there is something that comes paid for Plesk and free (or included) for cPanel. Just let us know about such example and we will be able to respond more specifically. Sometimes it works in quite the opposite way - afaik Lets Encrypt integration was paid for cPanel but free in Plesk. If you don't like bundles, or any addons, you are more than welcome to not purchase them and go with bare Plesk license. Actually that would be exactly the reason why that staff comes as an addon or bundle - to help those who needs them without disturbing those who don't.

As for charging for all languages together - would you really need all 32 languages together? Am I getting it right? That's definitely not very common case and that's why we didn't consider that before. But you are welcome to raise such request at Feature Suggestions: Top (1480 ideas) – Your Ideas for Plesk and if there will be more people willing to bundle all languages together, it can be considered probably.

 

[Sergey L]

Speaking of Docker now. You are absolutely correct saying that Plesk cannot deliver Docker for multitenant users. That is unfortunate Docker constraint coming from its security model.

However, it would have been quite an exaggeration to say the only need in Docker is to "provide something to someone". Actually we have a lot of users running Plesk in single-tenant mode, and they can easily enjoy running their services in Docker even at the same server w/o security risk coming from multi-tenancy. We are glad to make our singletenant users happy with Docker and we can admit we weren't able yet to make it work for multi-tenant users yet

For multi-tenant use, KuberDock for cPanel (by CloudLinux) might have been a solution for you, but they chose to shut it down recently.

Surely everyone is free to pick Racnher or another Docker management platform - we don't intend to substitute them at all, but only want to simplify Docker use in Plesk infrastructure and workflows.

 

[Ivan]

In the past, you offered Presence Builder with the highest edition. That product is EOL

As far as I know, Web Presence Builder is not going to be dropped. It is still supported. However, no further development is expected. So it is quite stable piece of Plesk.

 

[Sergey L]

How about changing the root or /home path for a particular subscription or customer? Is this now possible? What if I want some accounts to be stored in a different storage in a server? Can this be done now?

From our perspective, in Linux it can be easily done w/o change paths in Plesk - you just mount a storage into selected folder
Perhaps eventually we will let placing subscriptions in an arbitrary folder, but for now we see this as a rare need, which might be easily addressed by native Linux mount means

 

[Sergey L]

you can build your own texts too if you build your own language files

And here is a guide how to make your own language files for Plesk

 

[Sergey L]

Now on pricing policy

It seems everything you add to Plesk just increases it. This is the big and number one reason people stick and move to cPanel. One pricing, everything included. Not gimmicks.

Plesk has more flexible price than cPanel - there is a cheaper version of Plesk and more expensive versions. The most expensive Web Host edition would be $150 a year vs $200 in cPanel, plus there would be functions missing in cPanel. I don't think cPanel will have Git or Docker included at all, or WordPress Toolkit or CGroups Manager. And you can add more on top of it. While "more" costs extra - just compare it with the same cost in cPanel. When you have two calculations side by side - you can see which solutions delivers you more for the same price, or the same value for lower price. Just a math.

We will be more than happy to discuss if Plesk appears to be more expensive for some cases.

i.e. assuming you want to control resources per user and be able to pick different PHP versions per site (PHP5 for one and PHP7 for another) - you will be buying CloudLinux in cPanel. that's $14x12 = +$150/year; Assuming that you would want app installer, you will buy Fantastico DeLuxe for cPanel - +$164. So that's already $500+/year in cPanel but all 3 options (app installer, cgroups, multiple PHP) are included in Plesk's basic $150/year. So... is it really fair to see cPanel as "one price" and Plesk "just increases"?

Should you open applications.cpanel.com (somewhat similar to extensions/addons of Plesk) - would they all be free?

Looks like both panel have additional paid options, just selection might vary.

 

[Sergey L]

Multi-server, excellent, but not affordable. Really? $100 per server a year? Seems like a big piece of what the whole Plesk software costs per year. At that pricing, you are better just using something like Ansible or another automation tool.

We are more than happy if Ansible works for our users. But let me clarify here - Ansible is a perfect replacement for cpanel's multiserver, which is focused on centralized server settings management (afaik). Plesk Multiserver is a bit different and provides centralized account management - I am afraid Ansible cannot help with UI for that.
Did you mean copying accounts to different machines in the background? It might work only if machines are 100% identical

 

[Sergey L]

What you need to understand is that shared hosting as we know it is dead. The only reason its still working is because CloudLinux is solving some of the long problems, like security and resource charging.

We would agree here.

So unless Plesk plans to dockerize customers or start using namespaces and file system virtualization to cage users into their chroot environment's, you have no future because hosting shared companies have no future.

We do have certain plans to secure our future.

If you are interested to talk about dockerization, Plesk future, your other issues with Plesk, switching to Plesk, etc - we can arrange a call. PM me if interested

The newbie webite market is taken away by companies like Weebly, no solution with Pesk for that.

Partly true. Weebly, etc are active in partnering CPs. We have 4 different sitebuilder solutions partnering Plesk and Weebly is partnering cPanel as you know.

And the bigger customers need powerful resources and security for which shared hosting is not a solution either.

True. But shared hosting is only a part of Plesk business. Much larger part is single-tenant use of Plesk for managing sites of single organization in VPS.

 

[Sergey L]

One of the reasons some people moved to Plesk, was actually that software. No joke. I used that software with cPanel servers as well, and it was by far the most decent sitebuilder available (not great, but decent from all the garbage out).

We do agree - out of all hosted sitebuilders, Presence Builder was one of the best at that time. We still support it BTW.

As companies lose customers over to Weebly and Wix, not having something to offer users that have no idea about WordPress and just want a simple drag and drop interface was a strong selling with Plesk.

BoldGrid combines drag and drop experience with power of WordPress platform. We can expect a tighter integration soon.
Few other sitebuilders are coming through extensions.
Yet we believe that everyone shall focus on their area of competence, so future solutions will be integrated rather than built-in.

Providing rock solid email services, a great sitebuilder, a platform for easy and click launch applications, you get the idea. Whatever you are thinking Plesk team, think bigger.

You might be interested to checkout

• Kolab for email - it is great and it is more than just email.
• aforementioned BoldGrid, Web Business Builder by Yola and few more are coming as sitebuilders
• click and launch applications via built-in App Catalog (free, included) and WordPress Toolkit (included in top editions)

 

[orion]

Hi, and thank you for your feedback.
There were many topics brought into discussion, we will try to handle one at a time.

I am not sure what sort of branding customizations you lack comparing to cPanel. If you can show a particular need that you were unable to fulfill, then we can talk about the particular problem and perhaps plan an improvement

What I can see relevant to this problem now:
- you can disable a lot in Plesk UI through panel.ini file and if you don't like Extensions particularly you can disable extensions as a whole or individually
- you can also build your own view of Plesk on CSS and images level. Yes, you might need to adjust it with new versions of Plesk coming
- you can build your own texts too if you build your own language files. As this is requested only once every couple years, we don't have a public documentation probably - yet it is not really difficult for anyone who knows Plesk a little bit. A piece of advise - if you develop it as an extra language (rather than replace English locale), it will be tolerant to future updates as texts missing in your locale will be taken from en-US language.

You can checkout documentation for other options and let us know if you don't find something essential

Hello Sergey, thank you for the reply and all the information. It seems that indeed you can customize Plesk more today. I was talking mainly based on my past experience (years back) where the options mainly limited to the logo and a button or link and that was pretty much it. I started to explore my local installation, and it seems that indeed modifying themes and text seems possible now, as well build extensions or things on top of the GUI. Sure, you will have to merge changes with upgrades, but this applies to any software if you decide to go the customization route, manually code changes is usually what companies do when they have custom environments, plugins, services, integrations, etc. that all need to talk together. A control panel is just one more part fo that equation.

The links you pointed and after some reading have convinced me that Plesk is much more flexible today. So that is great news! I always loved Plesk documentation as its very well done, same is true for the translations, as someone that speaks three languages, I can say that Plesk has done by far the best job regarding quality and consistency.

 

[orion]

I am not sure if there is something that comes paid for Plesk and free (or included) for cPanel. Just let us know about such example and we will be able to respond more specifically. Sometimes it works in quite the opposite way - afaik Lets Encrypt integration was paid for cPanel but free in Plesk. If you don't like bundles, or any addons, you are more than welcome to not purchase them and go with bare Plesk license. Actually that would be exactly the reason why that staff comes as an addon or bundle - to help those who needs them without disturbing those who don't.

As for charging for all languages together - would you really need all 32 languages together? Am I getting it right? That's definitely not very common case and that's why we didn't consider that before. But you are welcome to raise such request at Feature Suggestions: Top (1480 ideas) – Your Ideas for Plesk and if there will be more people willing to bundle all languages together, it can be considered probably.

cPanel never charged for Let's Encrypt, at least I was never aware of that because I surely never paid for it on the cPanel servers that are around the DC. I don't remind cPanel charging for any new feature at all ever. They were implemented for everyone with an active license.

As for languages, of course, you will probably not need every single language. But assuming you are focusing on the European market alone, where Plesk seems strong, that alone requires more than one language, English, German, French, Spanish, Dutch, Italian, other....

I just think it's very odd to charge that much for extra languages. Maybe have packs of 3-5, etc. Or for individual languages charge less. You have to consider here that if a company has multiple servers, they would need to have this PER every server if they want consistency. While you can have one server with more languages and others with one, that is a management nightmare if you are segregating customers based on language (one server for Germans, another for Italian's, etc.).
One of the benefits from shared hosters is that they balance resources around by moving accounts based on load performance, disk usage, etc. At least I never considered did that. Servers are consistent in configuration, hardware, etc. They basically are all equal so the experience and services are the same regardless of where a customer was activated, migrated or transferred. This is probably what most companies also do, consistent configurations.

 

[orion]

Speaking of Docker now. You are absolutely correct saying that Plesk cannot deliver Docker for multitenant users. That is unfortunate Docker constraint coming from its security model.

However, it would have been quite an exaggeration to say the only need in Docker is to "provide something to someone". Actually we have a lot of users running Plesk in single-tenant mode, and they can easily enjoy running their services in Docker even at the same server w/o security risk coming from multi-tenancy. We are glad to make our singletenant users happy with Docker and we can admit we weren't able yet to make it work for multi-tenant users yet

For multi-tenant use, KuberDock for cPanel (by CloudLinux) might have been a solution for you, but they chose to shut it down recently.

Surely everyone is free to pick Racnher or another Docker management platform - we don't intend to substitute them at all, but only want to simplify Docker use in Plesk infrastructure and workflows.

I forgot about the single tenant mode. So you are right here. If you are using Plesk for your own sites and organization, (not providing access to others or selling services). Then yes, the feature is useful.
For the hosting edition or users in a multi-tenant environment. Not at all. And while containers like Docker are far more secure today than in the past, they are still not there to the point of production ready when it comes to security. So isolating containers groups with VM's is what most companies do today. This means one VM per server and containers running inside. So the only real Docker option for providers is the extra extension and like I said, it's limited to 1 external server which is rather odd as it means 1 customer per Plesk server.

You could argue here, that you can just provision services and not give user access to containers, but that is just the same story with a higher level access. You can't give users access to anything on containers, not even services, that includes no MySQL container or WordPress container, or anything. Not today at least, for at least some time, using an hypervisor with containers will be the way to go. Microsoft is doing it this way with hyper-v and so does everyone else, VmWare, Xen, KVM. Virtualization is actually perfect with containers and solves that problem altogether. But that again means multiple remote Docker servers per Plesk server if you want customers to be able to manage containers from his GUI side.

 

[orion]

As far as I know, Web Presence Builder is not going to be dropped. It is still supported. However, no further development is expected. So it is quite stable piece of Plesk.

That is not the information we received, and that is here on the forums as well. As far as I know, Presence Builder does not receive any features and is not developed anymore. Only bug fixes. And it has an EOL date already established.

You can't expect someone to purchase Plesk based on a feature that is going to be removed.
Its sad as its very decent actually, with a few more updates and features it could really be a very attractive selling point for Plesk.

 

[orion]

The weebly/sitebuilder users. Those that are not Drupal, Wordpress or Joomla users. They are not developers. I must sadly say that things like Weebly have drone a terrible impact on both cPanel, Plesk and most shared companies. Those users don’t know what PHP is, or MySQL, and they surely don’t want to use FTP. They don’t care about hosting, servers or the control panel. All they want is email and website. This is why those solutions work so well. And for my surprise, those people are willing to pay more money than someone just running WordPress. The regular developer and webmaster is not cheap, but very cheap. The person that starts building a site online for the first time, is saving so much money by not paying a web developer that is willing to pay 10 times as much as the regular person looking for shared hosting. Of course, something like those services will never replace a web designer or developer but that is not even the idea of those services. Those users don’t want a complex site.

I see you pointed out a solution based on WordPress but that is exactly what you don’t want for those users. Never give WordPress, PHP or MySQL to those newbie users. The reason you want static build HTML sites for them is very simple. It works, never has performance issues, never has security issues and its easier for entry users to the Internet to understand. They are not programmers but know how to create an .html file with Word. So the concept of HTML is not only better for them but also better for the hosting provider. Those customers are pure profit because they pay well, never have security issues (you can’t hack HTML), and they also never have performance issues, even on high traffic sites you don’t even notice them because nothing beats pure html with no database.

The only thing those people want is an extremely easy drag and drop builder and not having to code anything by hand. Time and money are the factors here for them. A visual builder is exactly what they want. That and having a huge selection of nice templates, colors and graphics. They want a website, a presence online, so the name Presence Builder was actually amazingly perfect. Who ever came up with that name on Odin/Plesk really knew the concept. Those people just want to be online and they are doing it themselves because they are internal artists, love to build things themselves or in most cases don’t have hundreds or thousands of dollars to pay a web developer so they go with something like Wix or Weebly.
The problem with offering or bundling those things is that they are just fancy affiliate programs. I don't feel comfortable as hosting provider to send customers to someone else. And the pricing also means you have to add costs on top of them which tends to be much higher than those users going directly with those services. You can't compete with those services unless you run your own SiteBuilder and have it a bit customized with unique services that those services can't provide or have no interested. In that case, the customer will stick with your company because you are providing a unique solution. Presence Builder was a nice product, while very simple it had a great future ahead, and then suddenly it was killed.

 

[orion]

About changing the storage path. The reason is to offer different or more complex hosting plans. Other control panel's can do this already.

Imagine the following scenario I post here as example:

You provide 10 email accounts but limit each account to 1 GB in size on X hosting plan.

You also have an email hosting plan mainly targeted for heavy email users (yes, email is actually another profit revenue source if you target small business correctly). But you provide 100 GB per account on that plan. And you want users on plan X to upgrade to that plan if email is their main concern.

The first one, email is included free or as part of a current web hosting package. It’s an all inclusive package, like most budget hosting companies are doing it for years, where everything is included for one simple price. The second one is per user, has better spam filters, or overall is a more complex, email dedicated solution.

It would make no sense to send customers to another control panel just for email. So can Plesk do it? In theory, yes. Technically it would be a bit more challenging.

The reason is that you don’t want to use the same storage for web users as for heavy email users. Website storage is expensive, high performance. Email storage is file storage, cheap, bulk.

So technically the solution here is to have a different storage path for the email plan. You can’t otherwise do it. That means having a different email storage or /home storage for customers on that plan.

And before you argue, you can just deploy a Plesk server for email, sure you can, but that is even more of a problem for those users. Now they have to log into one server for email, and another one for everything else. If you need to start deploying different Plesk servers for each service, then Plesk fails as control panel. The idea of a control panel is having a centralized interface so that customers can manage all the services.