@toor said: Swap Space: Permitted, but overuse (more than 25-30% of RAM) may lead to performance issues and warnings.
25% of what? Swap or Physical?
They don't want your swap to be over 25-30% of the RAM of your VPS plan. So if you have a 1GB VPS, the swap should not be larger than 250-300mb.
This is odd, i've never seen provider limit swap usage like this. How's this affect performance anyways?
Buy VPS with small RAM, add mucho swap, overload with applications so as to utilize the swap as RAM most of the time, kill node disk I/O. Profit. Although I don't expect that limiting swap to 25% will actually help disk I/O.
@toor said: Swap Space: Permitted, but overuse (more than 25-30% of RAM) may lead to performance issues and warnings.
25% of what? Swap or Physical?
They don't want your swap to be over 25-30% of the RAM of your VPS plan. So if you have a 1GB VPS, the swap should not be larger than 250-300mb.
This is odd, i've never seen provider limit swap usage like this. How's this affect performance anyways?
Buy VPS with small RAM, add mucho swap, overload with applications so as to utilize the swap as RAM most of the time, kill node disk I/O. Profit. Although I don't expect that limiting swap to 25% will actually help disk I/O.
@FrankZ said:
Buy VPS with small RAM, add mucho swap, overload with applications so as to utilize the swap as RAM most of the time, kill node disk I/O. Profit. Although I don't expect that limiting swap to 25% will actually help disk I/O.
More eloquently said as always. ;-)
Which is why you beat me to the punch. I take too much time to compose comments.
@toor said: Swap Space: Permitted, but overuse (more than 25-30% of RAM) may lead to performance issues and warnings.
25% of what? Swap or Physical?
They don't want your swap to be over 25-30% of the RAM of your VPS plan. So if you have a 1GB VPS, the swap should not be larger than 250-300mb.
This is odd, i've never seen provider limit swap usage like this. How's this affect performance anyways?
Buy VPS with small RAM, add mucho swap, overload with applications so as to utilize the swap as RAM most of the time, kill node disk I/O. Profit. Although I don't expect that limiting swap to 25% will actually help disk I/O.
If disk IO is a concern, I would opt for IOPS limitation. As for swap, I believe that zram is a better option even for lowend CPU.
@skhron said: As for swap, I believe that zram is a better option even for lowend CPU.
Then you'll end up hitting vCPU share limits due to the compression overhead (on low-end VPS), so it's a toss-up.
The provider writes strict limits as an excuse to delete customer VPS without warning.
VirmAche and GreenCloudVPS pioneered this practice.
It's well known that these VPS are only for idling and cannot be actually used.
@skhron said: As for swap, I believe that zram is a better option even for lowend CPU.
Then you'll end up hitting vCPU share limits due to the compression overhead (on low-end VPS), so it's a toss-up.
The provider writes strict limits as an excuse to delete customer VPS without warning.
VirmAche and GreenCloudVPS pioneered this practice.
It's well known that these VPS are only for idling and cannot be actually used.
That's too bad, I thought I liked my Charity Host VPS, but now it costs $25 to use any email related port and decreased CPU fair use sucks.
Also, it seems even idling could make you hit the vCPU share limits:
Discounted VPS – up to 25% of total CPU core capacity. Sustained CPU% is calculated from the hypervisor host.
NOTE: Depending on the compute workload, the hypervisor may register 10% to 35% more CPU usage than the VPS metrics itself due to virtualization overheads.
In the best case: 25% - 10% means the vCPU limit is really 15%.
In the worst case: 25% - 35% mean the vCPU limit is really -10%.
So if you are idling at 0%, the hypervisor could register your usage at 35%, meaning you are 10% past your limit lol
@skhron said: As for swap, I believe that zram is a better option even for lowend CPU.
Then you'll end up hitting vCPU share limits due to the compression overhead (on low-end VPS), so it's a toss-up.
The provider writes strict limits as an excuse to delete customer VPS without warning.
VirmAche and GreenCloudVPS pioneered this practice.
It's well known that these VPS are only for idling and cannot be actually used.
That's too bad, I thought I liked my Charity Host VPS, but now it costs $25 to use any email related port and decreased CPU fair use sucks.
Also, it seems even idling could make you hit the vCPU share limits:
Discounted VPS – up to 25% of total CPU core capacity. Sustained CPU% is calculated from the hypervisor host.
NOTE: Depending on the compute workload, the hypervisor may register 10% to 35% more CPU usage than the VPS metrics itself due to virtualization overheads.
In the best case: 25% - 10% means the vCPU limit is really 15%.
In the worst case: 25% - 35% mean the vCPU limit is really -10%.
So if you are idling at 0%, the hypervisor could register your usage at 35%, meaning you are 10% past your limit lol
Has the provider closed email related ports that you were using and asked you for $25 to reopen them?
Normally this type of TOS change, port closings, allows current customers to be grandfathered and only effects customer purchases after the TOS effective date.
This kind of change, counting hypervisor overhead & reducing fair share limits, shortly after purchase seems kind of shady to me.
I'd be interested to hear if this change actually becomes an issue for LESbians, or if the provider is just trying to cover his ass for abusers and continues to apply this TOS change fairly to the average user.
NOTE: This provider has not been active on LES since Dec 14th, 2024, so I don't expect a clarification from the provider.
@skhron said: As for swap, I believe that zram is a better option even for lowend CPU.
Then you'll end up hitting vCPU share limits due to the compression overhead (on low-end VPS), so it's a toss-up.
The provider writes strict limits as an excuse to delete customer VPS without warning.
VirmAche and GreenCloudVPS pioneered this practice.
It's well known that these VPS are only for idling and cannot be actually used.
That's too bad, I thought I liked my Charity Host VPS, but now it costs $25 to use any email related port and decreased CPU fair use sucks.
Also, it seems even idling could make you hit the vCPU share limits:
Discounted VPS – up to 25% of total CPU core capacity. Sustained CPU% is calculated from the hypervisor host.
NOTE: Depending on the compute workload, the hypervisor may register 10% to 35% more CPU usage than the VPS metrics itself due to virtualization overheads.
In the best case: 25% - 10% means the vCPU limit is really 15%.
In the worst case: 25% - 35% mean the vCPU limit is really -10%.
So if you are idling at 0%, the hypervisor could register your usage at 35%, meaning you are 10% past your limit lol
Has the provider closed email related ports that you were using and asked you for $25 to reopen them?
Normally this type of TOS change, port closings, allows current customers to be grandfathered and only effects customer purchases after the TOS effective date.
I haven't set up an MTA on that particular VPS yet so I will have to try it out. Yesterday 1/31/2025, I received an email addressed to me that the Acceptable Use Policy (AUP) has been updated, effective Friday, 31st January 2025, and that I should familiarize myself with these changes, so from that it sounded like all of it applies to a server I already purchased.
Title : Refactoring of VPS HDD Storage services and limitations
Dear [],
We have created new VPS HDD Storage plans to focus on maintaining stability, reliability and performance for all VPS HDD Storage customers.
Recently, we have had to take action on VPS HDD Storage service AUP violations. In order prevent these occurences we have refactored the VPS HDD Storage plans to accodomate all Storage and Backup use cases for VPS HDD Storage customers. These plans are intended for storage and backups, not for compute. The underlying host has IO and resource limitations that cannot accomodate excessive use of compute and IO for services on these VPS.
Due to this, we kindly ask that you focus the use of your VPS HDD Storage service for actual data storage and backups, and not use these as primary compute for services.
We have tried to keep your legacy resources as is, however, as we mentioned, when compute use is excessive, we have to act to protect your VPS as well as others on these hosts. Therefor we have had to align current VPS HDD Storage plans to the new model.
Apologies for any inconveniences.
We thank you for your understanding and cooperation.
Title : Refactoring of VPS HDD Storage services and limitations
...In order prevent these occurences we have refactored the VPS HDD Storage plans to accodomate all Storage and Backup use cases for VPS HDD Storage customers.
...
We have tried to keep your legacy resources as is, however, as we mentioned, when compute use is excessive, we have to act to protect your VPS as well as others on these hosts. Therefor we have had to align current VPS HDD Storage plans to the new model.
...
What does all that even mean? What is the new model? What does, accodomate (accommodate?) all Storage and Backup use cases for VPS HDD Storage customers mean? So legacy resources have, can, or will be changed?
I think they are going about this backwards. I think they need to identify the abusers (which it sounds like they can and have), warn them once, then if they continue, kick them to the curb with no refund for violating TOS/AUP. That way they don't have to keep tightening the reigns on all customers, just to try and control a few ***hole abusers.
Title : Refactoring of VPS HDD Storage services and limitations
...In order prevent these occurences we have refactored the VPS HDD Storage plans to accodomate all Storage and Backup use cases for VPS HDD Storage customers.
...
We have tried to keep your legacy resources as is, however, as we mentioned, when compute use is excessive, we have to act to protect your VPS as well as others on these hosts. Therefor we have had to align current VPS HDD Storage plans to the new model.
...
What does all that even mean? What is the new model? What does, accodomate (accommodate?) all Storage and Backup use cases for VPS HDD Storage customers mean? So legacy resources have, can, or will be changed?
I think they are going about this backwards. I think they need to identify the abusers (which it sounds like they can and have), warn them once, then if they continue, kick them to the curb with no refund for violating TOS/AUP. That way they don't have to keep tightening the reigns on all customers, just to try and control a few ***hole abusers.
Welp, they didnt give much context on what happens. Maybe not just few ass abuser, may be a lot abusers since they ran 1.88$/m 500G HDD promotion last december on OGF. I believe this attract a lot of abusers on their HDD node
Comments
Buy VPS with small RAM, add mucho swap, overload with applications so as to utilize the swap as RAM most of the time, kill node disk I/O. Profit. Although I don't expect that limiting swap to 25% will actually help disk I/O.
For staff assistance or support issues please use the helpdesk ticket system at https://support.lowendspirit.com/index.php?a=add
More eloquently said as always. ;-)
Which is why you beat me to the punch. I take too much time to compose comments.
For staff assistance or support issues please use the helpdesk ticket system at https://support.lowendspirit.com/index.php?a=add
If disk IO is a concern, I would opt for IOPS limitation. As for swap, I believe that zram is a better option even for lowend CPU.
Check our KVM VPS (flags are clickable): 🇵🇱 VPS 🇸🇪 VPS 🇵🇱🇸🇪 VDS | Looking glass: 🇵🇱 🇸🇪
Then you'll end up hitting vCPU share limits due to the compression overhead (on low-end VPS), so it's a toss-up.
However its performance is higher and in case of abuse it is enough to throttle only CPU.
Check our KVM VPS (flags are clickable): 🇵🇱 VPS 🇸🇪 VPS 🇵🇱🇸🇪 VDS | Looking glass: 🇵🇱 🇸🇪
The provider writes strict limits as an excuse to delete customer VPS without warning.
VirmAche and GreenCloudVPS pioneered this practice.
It's well known that these VPS are only for idling and cannot be actually used.
No hostname left!
That's too bad, I thought I liked my Charity Host VPS, but now it costs $25 to use any email related port and decreased CPU fair use sucks.
Also, it seems even idling could make you hit the vCPU share limits:
15%.-10%.So if you are idling at 0%, the hypervisor could register your usage at 35%, meaning you are 10% past your limit lol
I ❤ servers.
Has the provider closed email related ports that you were using and asked you for $25 to reopen them?
Normally this type of TOS change, port closings, allows current customers to be grandfathered and only effects customer purchases after the TOS effective date.
This kind of change, counting hypervisor overhead & reducing fair share limits, shortly after purchase seems kind of shady to me.
I'd be interested to hear if this change actually becomes an issue for LESbians, or if the provider is just trying to cover his ass for abusers and continues to apply this TOS change fairly to the average user.
NOTE: This provider has not been active on LES since Dec 14th, 2024, so I don't expect a clarification from the provider.
For staff assistance or support issues please use the helpdesk ticket system at https://support.lowendspirit.com/index.php?a=add
I haven't set up an MTA on that particular VPS yet so I will have to try it out. Yesterday 1/31/2025, I received an email addressed to me that the Acceptable Use Policy (AUP) has been updated, effective Friday, 31st January 2025, and that I should familiarize myself with these changes, so from that it sounded like all of it applies to a server I already purchased.
I ❤ servers.
AUP has changed again:
https://charityhost.org/acceptable-use-policy/
Archive: https://web.archive.org/web/20250202194139/https://charityhost.org/acceptable-use-policy/
Diff (Red = Old, Green = New)
… after we stirred up drama on OGF.
No hostname left!
@CharityHost_org is Prem.
I understand having issues with change in AUS but I think it wil better their services in the long run.
I got an email from charityhosting
Title : Refactoring of VPS HDD Storage services and limitations
Dear [],
We have created new VPS HDD Storage plans to focus on maintaining stability, reliability and performance for all VPS HDD Storage customers.
Recently, we have had to take action on VPS HDD Storage service AUP violations. In order prevent these occurences we have refactored the VPS HDD Storage plans to accodomate all Storage and Backup use cases for VPS HDD Storage customers. These plans are intended for storage and backups, not for compute. The underlying host has IO and resource limitations that cannot accomodate excessive use of compute and IO for services on these VPS.
Due to this, we kindly ask that you focus the use of your VPS HDD Storage service for actual data storage and backups, and not use these as primary compute for services.
We have tried to keep your legacy resources as is, however, as we mentioned, when compute use is excessive, we have to act to protect your VPS as well as others on these hosts. Therefor we have had to align current VPS HDD Storage plans to the new model.
Apologies for any inconveniences.
We thank you for your understanding and cooperation.
What does all that even mean? What is the new model? What does, accodomate (accommodate?) all Storage and Backup use cases for VPS HDD Storage customers mean? So legacy resources have, can, or will be changed?
I think they are going about this backwards. I think they need to identify the abusers (which it sounds like they can and have), warn them once, then if they continue, kick them to the curb with no refund for violating TOS/AUP. That way they don't have to keep tightening the reigns on all customers, just to try and control a few ***hole abusers.
I ❤ servers.
Welp, they didnt give much context on what happens. Maybe not just few ass abuser, may be a lot abusers since they ran 1.88$/m 500G HDD promotion last december on OGF. I believe this attract a lot of abusers on their HDD node