Why do hosts avoid stacking their LET-offered packages?

Mumbly

I myself rarely use a really big VPS, so I am not affected by this, but I see it on a regular basis. Why do so many hosts avoid stacking their offered packages?
Is it because of the manual work involved? Or maybe because people who require this actually use their resources? What's the reason?

From where I stand, this could often actually be profitable for a host, since stacked VPSes usually don't include IPs. IPs are often one of the main fixed costs, and if a host can save some by stacking VPSes but charge the same price... what's not to like for the host?
I understand it with exclusive, super limited offers, but for everything else… what's the reason for not doing this - merging the resources of two VPSes into one, but with only one IP, at the same price as two VPSes?
From what I see, people who ask if hosts allow VPS stacking are usually interested in merged resources but are willing to take only one IP.

Genuinely curious. Maybe some hosts can chime in?

NotFoundException

You can migrate and balance smaller VMs more easily than with a few big VMs. Usually you don’t expect every customer to use the resources 100% all the time, with smaller VMs you can react better if someone does.

noloudelou

I guess because the profit is low and they want to attract more new users

layer7

Hi,

most offers are economically not sustainable if really used. If a customer would use too much, the hoster would actually loose money, not make money. Keyword here is: Overselling

So the strategy with those small packages is, that you usually can not really do something serious with them. They are good for a VPN / Proxy maybe some lightweight file/web/what ever server. But nothing real where applications which requires CPU/RAM are able to run.

Thats the reason why you see them for 1-2 EUR and even less.

Now if a hoster would allow stacking, he would actually contradict this whole system that enables this small prices. If you can combine multiple small servers for 1-2 EUR or even less ( there are 70 cent servers out there ^^; i think ) you can actually create one big server with multiple cores/ram/disk that is actually ready to run something real on it ( CPU/RAM/Networkwise ). But thats exactly what the hoster does NOT want. He/she wants you to pay money, but not use it -> Idle for the win!

Otherwise this is simply not working. I know some customers do not do any math if an offer can actually work out or is doomed to be deadpooled while the question is just "when" and not "if". But checkout what hardware cost. Colocation cost. Traffic cost. Then take the selling price of the hoster and math how much instances the hoster would have to install on this hardware just to cover this cost ( no profit ). You will fastly see that this is not working at all ( if a majority is not mostly idling ).

This can work as long as the instances are just non sense enough. ( Big disks, no traffic = not suitable for real fileservers, small CPU/RAM/disk, very small CPU OR RAM OR Disk like 256 MB RAM or so ). This way the hoster raise the chance that its just idling ( as you can anyway not do something real with it ).

But if the hoster allows stacking for the same price... he will actually just make sure to deadpool earlier.

mans_xd

@layer7 said:
Hi,

most offers are economically not sustainable if really used. If a customer would use too much, the hoster would actually loose money, not make money. Keyword here is: Overselling

So the strategy with those small packages is, that you usually can not really do something serious with them. They are good for a VPN / Proxy maybe some lightweight file/web/what ever server. But nothing real where applications which requires CPU/RAM are able to run.

Thats the reason why you see them for 1-2 EUR and even less.

Now if a hoster would allow stacking, he would actually contradict this whole system that enables this small prices. If you can combine multiple small servers for 1-2 EUR or even less ( there are 70 cent servers out there ^^; i think ) you can actually create one big server with multiple cores/ram/disk that is actually ready to run something real on it ( CPU/RAM/Networkwise ). But thats exactly what the hoster does NOT want. He/she wants you to pay money, but not use it -> Idle for the win!

Otherwise this is simply not working. I know some customers do not do any math if an offer can actually work out or is doomed to be deadpooled while the question is just "when" and not "if". But checkout what hardware cost. Colocation cost. Traffic cost. Then take the selling price of the hoster and math how much instances the hoster would have to install on this hardware just to cover this cost ( no profit ). You will fastly see that this is not working at all ( if a majority is not mostly idling ).

This can work as long as the instances are just non sense enough. ( Big disks, no traffic = not suitable for real fileservers, small CPU/RAM/disk, very small CPU OR RAM OR Disk like 256 MB RAM or so ). This way the hoster raise the chance that its just idling ( as you can anyway not do something real with it ).

But if the hoster allows stacking for the same price... he will actually just make sure to deadpool earlier.

i get what you mean if i do simple Calculation 8GR-4VC 120 nvme Amd
in layer 7 with 6€ but if try to double resources pricing are going to be 14,19 EUR

layer7 still layer7

oloke

@layer7 said:
This way the hoster raise the chance that its just idling ( as you can anyway not do something real with it ).

So the hosts are purposefully gatekeeping me from not idling my servers?

Makes sense now, I started to think it was on me for idling them so hard

mans_xd

@oloke said:

show previous quotes

@layer7 said:
This way the hoster raise the chance that its just idling ( as you can anyway not do something real with it ).

So the hosts are purposefully gatekeeping me from not idling my servers?

Makes sense now, I started to think it was on me for idling them so hard

oliver reveals secrets

rpqu

@layer7 said:
Idle for the win!

This is a good for a banner