Question Massive price increase for reseller licenses?

[Alejandro Oro Vojacek]

Open Lite Speed hands down beat the hell out of nginx + apache in the way they are setup in Plesk.
Same identical server that hold 200 domains on Plesk (most of them on NGINX) were hitting 50% of the CPU on a Ryzen 5 5600.
Same identical server with those 200 domains on Webuzo with Open Lite Speed and I get less than 20% of CPU usage.

This is because Plesk made it so difficult to use native nginx, even when you have EVERYTHING setup properly, there will be sites that ends up using the proxy. So you are never 100% in control of what happens.

I already tried going domain by domain and pushing an NGINX only setup. Also tried making sure that the subscriptions attached to the user were not using any proxy, but Plesk insist on enabling the proxy from time to time and I completely lost any hope they are going to fix it.

With Open Lite Speed it's just Open Lite Speed, there is no margin of error to call upon Apache or any other webserver and the way it handles the LSPHP processes is way more efficient. I won't go into details. Suffice to say, at the end of the day, those same servers are more efficient and allows my customers to run more snappy installations due to less overhead on the background.

 

[seqoi]

Well, I just canceled all remaining Plesk subscriptions as a direct countermeasure of their (in my case) completely unjustified price increases. My next payment was in two months. I have already absorbed two (or three) price increases over the years of using Plesk. And I haven't seen or "felt" any improvement they are speaking about as a justification for their price increases.

The fact that they used the same newsletter template from the past year to send me such notice made me feel like a lowlife and I took it personally even though I rarely mix emotions with business.

If the sign of them not being able to invest in a new design for a friggin newsletter that will upset 99% of their userbase - is not pure evidence of questionable tactic of "saving" and adding burden on their existing customers and a telltale story of what we can expect from them in the future - I don't know what is.

I have two months to test Webuzo or ApisCP. These filtered out as favorites (for my line of work).

Goodbye boys. Farewell.

 

[seqoi]

And.... I am back

Back to Plesk. Just ordered a new license, and a new server and am about to manually update MariaDB to 11.4.

TLDR - yes there are many good options and cheaper alternatives to Plesk, such as ApisCP. I tried many other suggestions in this thread.

If you are someone running a small business and you need a secure and cheap panel with constant updates take ApisCP. NOTE You need to know the Linux operating system and understand it. It is NOT for the newbies. You need to understand the terminology and best practices.

This brings me to a Plesk. There's one Plesk advantage over the other panels for our case (I emphasize our/mine).

Yes, my complaints are still valid. But Plesk workflow and UI are so mature that I can allow someone who doesn't know Linux to be able to operate a server via the Plesk interface - without them knowing too much about intricate details of OS. It is on the verge of being suicidal masochist to allow someone who doesn't know Linux OS to control your server via Plesk - however, at a time when I am out of the office and I need some of my colleagues to do usual maintenance Plesk is much more matured product and I can trust and allow some of my team to handle the small stuff. That is not happening with ApisCP.

So really yes Plesk will be becoming a shitstorm unless they slow down. After the email fiasco introduced a month ago (they fixed it by now) I suggest they at least bring back early adopter updates.

So is Plesk bad with all the ongoing features no one asks for and price hikes? No. And Yes. It depends what you need for your business.

I will argue it's still way better than a cPanel (for me).

+ On this forum You get to be in contact with their staff personnel which also happens to be a super nice woman in every possible and I like to see women in this field of business - I hope I am not stepping on any toes or politically sensitive snowflakes. I simply love to see ladies here. Nothing to be upset about.

 

[tkalfaoglu]

For a small business like ours, the prices hikes are too much - I'm considering some big changes

 

[trialotto]

And.... I am back

Back to Plesk. Just ordered a new license, and a new server and am about to manually update MariaDB to 11.4.

TLDR - yes there are many good options and cheaper alternatives to Plesk, such as ApisCP. I tried many other suggestions in this thread.

If you are someone running a small business and you need a secure and cheap panel with constant updates take ApisCP. NOTE You need to know the Linux operating system and understand it. It is NOT for the newbies. You need to understand the terminology and best practices.

This brings me to a Plesk. There's one Plesk advantage over the other panels for our case (I emphasize our/mine).

Yes, my complaints are still valid. But Plesk workflow and UI are so mature that I can allow someone who doesn't know Linux to be able to operate a server via the Plesk interface - without them knowing too much about intricate details of OS. It is on the verge of being suicidal masochist to allow someone who doesn't know Linux OS to control your server via Plesk - however, at a time when I am out of the office and I need some of my colleagues to do usual maintenance Plesk is much more matured product and I can trust and allow some of my team to handle the small stuff. That is not happening with ApisCP.

So really yes Plesk will be becoming a shitstorm unless they slow down. After the email fiasco introduced a month ago (they fixed it by now) I suggest they at least bring back early adopter updates.

So is Plesk bad with all the ongoing features no one asks for and price hikes? No. And Yes. It depends what you need for your business.

I will argue it's still way better than a cPanel (for me).

+ On this forum You get to be in contact with their staff personnel which also happens to be a super nice woman in every possible and I like to see women in this field of business - I hope I am not stepping on any toes or politically sensitive snowflakes. I simply love to see ladies here. Nothing to be upset about.

Absolutely valid points ...... both the positive and negative points made.

However, I really appreciate your contribution due to the fact that is different from the "usual" contributions - different in a positive way.

Thank you!


PS Sure, Plesk Team should focus on gradual implementations of new features or improvements of old features, but that is another story ..... and it is good that that story is (again) brought to the attention of Plesk Team.

 

[trialotto]

Open Lite Speed hands down beat the hell out of nginx + apache in the way they are setup in Plesk.
Same identical server that hold 200 domains on Plesk (most of them on NGINX) were hitting 50% of the CPU on a Ryzen 5 5600.
Same identical server with those 200 domains on Webuzo with Open Lite Speed and I get less than 20% of CPU usage.

This is because Plesk made it so difficult to use native nginx, even when you have EVERYTHING setup properly, there will be sites that ends up using the proxy. So you are never 100% in control of what happens.

I already tried going domain by domain and pushing an NGINX only setup. Also tried making sure that the subscriptions attached to the user were not using any proxy, but Plesk insist on enabling the proxy from time to time and I completely lost any hope they are going to fix it.

With Open Lite Speed it's just Open Lite Speed, there is no margin of error to call upon Apache or any other webserver and the way it handles the LSPHP processes is way more efficient. I won't go into details. Suffice to say, at the end of the day, those same servers are more efficient and allows my customers to run more snappy installations due to less overhead on the background.

I am sorry, but this is comparing apples with pears.

And that is exactly the promise of Open Lite Speed, as demonstrated on their own page : a comparison of "speed" (i.e. performance) of WordPress sites on Open Lite Speed (LSCache) with Nginx (FastCGI cache) and Apache (W3 Total Cache).

I am not sure whether you compare 200 domains with WP instances or not.

However, the comparison of various alternatives with (very) specific caching mechanisms is very lopsided by nature.

Each and every caching mechanism is designed to serve a specific purpose.

Nginx is to some high degree different, in the sense that it is provided with a huge amount of "building blocks", allowing to serve multiple caching purposes.

Nevertheless, Nginx is born as a reverse proxy server, not as a caching mechanism.

So yes, there is the huge difference to start with.

In addition, when combining "building blocks" with Nginx, one still serves one or a couple of purposes - for instance, allowing caching modules with Nginx will result in a reverse proxy server with caching functionality.

Again, the primary objective is the reverse proxy server and the secondary objective is caching on the proxy server level.

This is primary + secondary objective combined is more or less the same to Open Lite Speed, with this difference that Nginx - the reverse proxy server itself - is more controllable and that the Nginx modular system is more flexible and can be augmented with each and every module one likes to add.

In essence, Open Lite Speed serves a specific purpose and is limited to that purpose - it is as good as it is programmed.

To be fair, Nginx might be less limited and the Nginx core might be excellently coded and the Nginx system might be more flexible, it still is limited.

In fact, Nginx is limited by the extent to which the implementation has been done properly.

More importantly, Nginx is often limited by the fact that the vast combinations in the setup of core and modules can create a performance pitfall - for most people, it is not clear how and when specific combinations will result in issues.

This is exactly the problem with Plesk.

Plesk has been configured in such a way that the default Nginx setup can never be optimal when activating Nginx based caching.

Stated differently, this well-known problem has been present for years and cannot be solved by tweaking settings in Plesk Panel or by adding custom Nginx config files on the server - one really has to create custom Nginx templates (and thanks to the nature of Plesk, the presence of these custom templates will always create issues of both minor and major degree).


So yes, one can conclude that Plesk with Nginx based caching will be outperformed frequently (but not always) by other caching solutions.

However, it still is a comparison of apples and pears.

Any pure Nginx configuration - without the burden of Plesk based default Nginx config - will always outperform all other caching solutions.

The problem is only that one needs to define "outperformance" - speed? resource usage? memory usage? request performance?

A proper and neutral definition of "outperformance" cannot be provided, in my humble opinion.

After all, optimal caching solutions should be dynamical - each every solution can only be optimal when being optimized for that one specific purpose that is the objective, with optimization being a dynamical process (and not static).

No site is the same and even situations to be optimized will change each and every second for individual sites.

Having a Ferrari, Porche or Lamborghini that can go fast does not imply that one will drive fast continuously, right?

Having an optimal caching solution that is "good on average" is often the best one can do.

Sure, Nginx can be programmed to dynamically adjust to each and every situation ...... in theory and practice ....... but should I care about each and every situation that can or might occur and that requires a (small or big) tweak in caching settings?

No.


If Open Lite Speed works for you, then that is good.

If Plesk with Nginx based caching does not work for you ......... well, that is predictable - this is an issue for almost every Plesk Panel owner.

I sincerely hope that Plesk Team will have a look at this Nginx related challenge soon - fingers crossed.


Kind regards......

 

[Alejandro Oro Vojacek]

It is not loopsided and wrong if you literally move an entire server from one system to the other.

I did this with our last server, about 232 domains that I had to move from Plesk to Webuzo (Apache + NGINX on Plesk to Open Lite Speed on Webuzo).

After I manually moved all the domains by exporting/importing (since webuzo does not have a Plesk importer) the load on the new server was considerably lower. I can imagine the complexity of Plesk and the simplicity of Webuzo contributed to this factor.

And since I use a brand new identical server for the destination, the comparison is STILL VALID.

Since my relationship with Plesk finished with that last server, I can truly say I'm happy that I've "got out".

 

[trialotto]

It is not loopsided and wrong if you literally move an entire server from one system to the other.

I did this with our last server, about 232 domains that I had to move from Plesk to Webuzo (Apache + NGINX on Plesk to Open Lite Speed on Webuzo).

After I manually moved all the domains by exporting/importing (since webuzo does not have a Plesk importer) the load on the new server was considerably lower. I can imagine the complexity of Plesk and the simplicity of Webuzo contributed to this factor.

And since I use a brand new identical server for the destination, the comparison is STILL VALID.

Since my relationship with Plesk finished with that last server, I can truly say I'm happy that I've "got out".

Again, Plesk is not able to provide an optimal Nginx based caching mechanism, as has been explained.

One cannot compare with Open Lite Speed on Webuzo.

In addition, the major bottleneck in Apache + Nginx setups on Plesk is still Apache, not Nginx.

Plesk itself is not really complex, it is a combination of "packages" that are tuned to work well with each other - this can go with a performance penalty.


I have done my portion of testing of servers and setups, but it always turns out to be a simple line of reasoning : do I have an issue? do I have a solution? do I want the solution? do I need the solution?

In essence, issues can be identified easily and solutions can be implemented ........ and in most cases, the solution is not even needed (and aggravates the issue or even creates another set of issues).

The advantage of Plesk is still that it is "one general package" (and to be fair, with some features that are not needed or requested).


All of the above does not really matter.

In fact, one should always choose the solution that works for the person in question - it might be a matter of opinion or related to a specific purpose.


Kind regards....

 

[seqoi]

It is not loopsided and wrong if you literally move an entire server from one system to the other.

I did this with our last server, about 232 domains that I had to move from Plesk to Webuzo (Apache + NGINX on Plesk to Open Lite Speed on Webuzo).

After I manually moved all the domains by exporting/importing (since webuzo does not have a Plesk importer) the load on the new server was considerably lower. I can imagine the complexity of Plesk and the simplicity of Webuzo contributed to this factor.

And since I use a brand new identical server for the destination, the comparison is STILL VALID.

Since my relationship with Plesk finished with that last server, I can truly say I'm happy that I've "got out".

I think what trialotto wants to say is that Plesk out of the box isn't tuned to perform in the way you see on Webuzo with Open Litespeed out of the box. And one should configure the server based on specific requirements. There's no one setup fits all scenarios.

I understand your point, because most of my customers run WordPress. And Lite Speed or Open Litespeed from my experience are preconfigured out of the box so to speak (in the lack of finding a better word) to perform and to give Wordpress users better results like without messing a lot with intricate Apache details. My assumption is that this is what you see and experience. You installed Webuzo and everything performs better. In that sense, Open Litespeed or Litespeed is much better. There's no and there should be no denial of your experience from that standpoint. Like, if it works then it works.

But this doesn't mean Plesk couldn't deliver the same if not better configuration and performance if tuned properly (emphasize on tuned properly).

One can argue that even well-tuned Apache (without NGINX) can outperform Open Litespeed and NGINX in certain situations.

 

[Alejandro Oro Vojacek]

Well I've been working with Plesk for the past 7 years. And I was never able to optimize it/tweak it the way I like it.
On Webuzo it worked great out of the box and even doing small optimizations are easier to acomplish since the webserver setup is much simpler.
Provided Open Lite Speed includes the LSCache which is super easy to install and configure.

Both situations leads to much better performance overall and the system is so much easier to tweak to.

So, aside from performance, Plesk has a lot of other issues included.

Here are some:

• Complexity of the webserver stack
• Complexity of the system in general
• Too much stuff that you cannot take out
• Inhability to install Plesk for specific tasks
• You cannot remove specific modules, it's you take it all or you leave it.
• Inability to completely remove Apache, which makes NGINX underperform most of the time
• Insistence on Plesk on using Apache as default, even when configured in the subscription
• Too complex subscription level/customer setup compared to other systems
• WP management is a pain in the butt, most notably because of how slow it has become

This and the obvious reason of the license cost prompted me to step away from Plesk. I could say a lot of things regarding Webuzo but the modularity of the system and the possibility of you installing things as "Apps" is the best feature of them all.
Want a webserver system without DNS, Mail services or anything else in there, you can do it. You want a strictly focused mail server, you can do it. You want an all-around server, you can do it. You want Apache, NGINX or Open Lite Speed only stack? you can do it.

For a small webhosting company, being able to tweak the control panel is paramount and without all those options, Plesk is a nuisanse to deal with. Let's not even talk about the backups and the extra functionality.

Everything is a subscription to them. By the time you finished adding everything that you liked, you are probably paying a fortune in licenses and your system is already too slow to performance as you want.

 

[tkalfaoglu]

It is not loopsided and wrong if you literally move an entire server from one system to the other.

I did this with our last server, about 232 domains that I had to move from Plesk to Webuzo (Apache + NGINX on Plesk to Open Lite Speed on Webuzo).

Why did you go with Webuzo? It seems as expensive as PLESK?

 

[Alejandro Oro Vojacek]

Why did you go with Webuzo? It seems as expensive as PLESK?

What?. Small license for 1 user it's 8 a month and I can host unlimited domains on a single account. Up to 15 user accounts goes to $15 a month (this already allowed me to maximize several servers with more than 300 domains each with just 14 customers each) and the full license goes to $25 all unlimited.

How is this "as expensive as Plesk" ?.

Plus the smaller Plesk licenses are completely USELESS. The intermediate Plesk license up to 30-domains?!!. Give me a break.

 

[tkalfaoglu]

What?. Small license for 1 user it's 8 a month and I can host unlimited domains on a single account. Up to 15 user accounts goes to $15 a month (this already allowed me to maximize several servers with more than 300 domains each with just 14 customers each) and the full license goes to $25 all unlimited.

How is this "as expensive as Plesk" ?.

Plus the smaller Plesk licenses are completely USELESS. The intermediate Plesk license up to 30-domains?!!. Give me a break.

Ohh when you mentioned two hundred domains, I assumed you needed 100+ users as well.

 

[Alejandro Oro Vojacek]

I can see why got you confused.
Here is where Plesk got it all wrong. Instead of limiting per-customer they limit the intermediate license per-domain, which renders them completely USELESS to even the most basic of uses.
Webuzo limits per-customer, not per domain. So most of my servers holds 14 or less customers but since I offered plans for small to medium web agencies, I can allocate those customers in a specific server and hold 10 to 50 domains each.

You can pretty much guess why this is a much better and smart move for small companies such as mine. And you can also see why Plesk is so dam greedy.