New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Again, for this kind of rule, it should be highlighted in the package listing to avoid confusion. That's the most proper way of doing it.
There is a "Question" section right under your package listing. Every important thing is answered and clarified over there but this rule that is more important than each of them is in ToS instead. Why not clarify it there?
We all well know not everyone will read the ToS. It's a very powerful defense to blame people who don't read the ToS. But the whole thing is simply not that simple.
I asked another reputable provider once how their policy is. The answer was like a consumer with a 4-core-plan would expect: "Stay under a load of 4 and you are fine.". Anything else may be understandable for a techie, but not for a normal buyer.
When you buy the package/service you have to check a box that you have read and will use the service according to the terms of service. If you don't read and just check it to be able to order then it is your problem. You have agreed to the terms when signing up.
You CAN NOT hide behind "no one ever reads the terms of service".
Does the German law protect customers from themself?
TO be fair the German law doesn't even apply to this situation. They're using the law totally out of context, 100%.
EDIT: Also no where does it say you can't use your "4" cores. It says you can't use 4 cores the entire darn day 24/7.
When you signup with a service you agree to all the terms they have set to you and you must comply with the laws that the company resides in. Just because you're in another country doesn't mean you're exempt from anything.
I'm like lolwut over your post.
It's not like you're gonna get a lawyer to sue a company across the pond. That'd be more money than what the service is worth.
Please do find a factual link to the law that you're talking about.
What I was trying to say is if you want to do your best to explain your policies upfront. Why not mention this important policy both in ToS and in the package listing?
This policy definitely alter if one will order a package or not.
The whole process is like this now: read package listing, read Q&A, decide to buy, choose location, enter server name, choose billing cycle, enter name, enter address, enter password, enter billing method, read ToS*, see the rule --> decide not to buy
(* or didn't read ToS, start moving vps, start using for a few days, get warning ticket -> understand the rule -> decide to move out)
Instead, why not let the process be like this: read package listing, see the rule --> decide not to buy
Why not save other people's time? because you don't care or something complicated?
I think I've said this before....
"Load" on Linux isn't just CPU time, it also includes disk I/O wait.
So if someone else hammers the disk, my I/O suffers and my load goes up.
Just sayin' (again).
Wut? Go get some sleep or something.
Suggestion for future possible business:
1. Read package
2. Read ToS
3. Ask questions if something in ToS is unclear
We have VPS range, where you will never be suspended for CPU abuse. PM me for more details.
@Radi, is funny "will never be suspended" vs your signature
Just use Vultr, all the CPU you want.
The Vultr marketing team is here...
Thanks. Lesson learned.
Many customers still don't and never will.
Good work.
I will only go for big names from now on. So long.
Go educate yourself. Why write a post that stinks with ignorance and arrogance, when someone told you about something that actually exists?
I think you live in a bubble that needs burst.
I need to clarify something I said earlier. Our AUP doesn't actually say, "You can't use 4 cores 24/7." I'll just copy/paste it here since several people have commented without actually reading it:
"As in all virtual server environments, CPU cores and disk I/O are shared among multiple VPSs. If your VPS is consistently maxing out one full core or more and impacting the performance of other client servers sharing the same host node, we may ask you to reduce your usage. Load may be generated by CPU usage and/or I/O. Disruptive load may result in a reboot, shutdown, and/or suspension of the VPS regardless of the time lapse involved."
You'll notice there's an and in there between maxing out one full core or more and impacting the performance of other client servers. That means both conditions have to be met before we take any action. As I indicated above, we only do so when other people are being impacted, not simply because your individual load is high. When this thread was originally opened, our policy was worded in a more strict manner. If you're able to use 4 cores 24/7 on a node without much other activity, we may never say a word to you. However, there may be times where you have to limit your usage in order to share. That seems to me to be common sense as far as virtual servers go. If you can't operate in a shared CPU environment, you need some form of dedicated server. Taking a quick survey of the TOS of some "big names", I see statements regarding warning, limiting, suspending, terminating, etc in cases of "excessive" CPU resource usage. True, we are a bit more specific and maybe more active about handling this sort of thing than others, but I'm not seeing any company out of the four I briefly reviewed saying you can just do whatever you want with all of your CPU access. You may have had that experience at another host (just like a lot of people have that experience with us), but that particular issue is still addressed in the TOS of our competitors, meaning they care to some degree.
Anyway, @joe0528 - The only reason you're complaining here in the first place is you agreed to something without reading it and didn't like the consequences. That doesn't necessarily render your feedback invalid, but it does make this discussion somewhat petty in nature. I don't really care what laws may or may not apply in Germany...no one can place an order on our website without checking a box saying he or she has both read and agreed to our TOS. The "but no one reads it" excuse isn't much ground to stand on in that light.
Because you can have 6x2TB disks and 64GB and split them to say 32-64 clients and have it dedicated.
How the hell will you dedicate 4cores per customer? 128-256 cores not threads? Show me a cpu that can do that and won't send costs sky high.
So what any provider out there is doing is considering the fact most ppl won't use all their cores 100% 24/7 so they can overprovision. For that reason they put in the AUP that rule. You can use the cpu as long as you don't affect other clients. As Nick said as long as that doesn't happen they don't even enforce the AUP.
IF you want a vps with 4cores you can use at 4 load aka 100% then I bet you could find one but the price would be multiple times what you pay atm. A usual xeon you can find on dedicated servers will have 4-8 cores, so 1-2 dedicated customers with your kind of thinking.
I doubt the law works exactly as you have it pictured and in any sane country you will lose the battle at court.
disclaimer: that's my opinion as a user and customer.
We have a similar policy, however, locking cores is not allowed. This happens in cases of cryptocurency mining or other cases of "grid computing" which are forbidden on all our plans.
Also, the 100% load clause applies like this:
1. For budget plans, you can never exceed the maximum load (as in the example above, 4 for a 4 vcpu plan). You can stay at exactly 100% but never go above, which it means you should consider spikes in your apps and use conservative settings.
2. For regular (pro) plans, excluding budget brands, such as overzold, xenpower, you can go over 100% for limited time if needed, such as a reddit of your page, a sudden jump in visitors, compilations and streaming demand, etc. We have the extra capacity for this.
3. For Biz plans, you can often and for a long time go over the limit, however, more than 200% is not accepted unless in short bursts, like half an hour or so, over 1k% is not allowed, period. That usually means hacking or major malfunction which could lead to data loss so we shutdown and give you a notice.
I just want to say something here quickly about this topic as it often makes me want to smack people through the screen. ESPECIALLY with low end providers, services are generally sold as FAIR SHARE. This means that you are granted access to a larger amount of resources than you may be given access to if you were going to purchase dedicated resources for a lower price with the understanding that you are sharing these resources with others and you 'shouldn't be a dick'.
99.98756% of the providers on this forum (Low End) offer fair-share resources, especially when it comes to CPU, so you should expect when buying any service from an offer on LET that the resources will be shared and that they are NOT dedicated. Now this is what is common, not what every provider subscribes to. For example, if you choose a provider like Digital ocean I think they give you full access to the cores your given, but most other smaller ventures are not going to do this as it is not economically feasible. So if you are trying to use a service like Digital Ocean as an example of what to expect in the low end, DO NOT as this is not the 'norm' but the exception to it.
You will know when you are paying for dedicated resources as it will cost you a lot more than $7.00 a month and will often be as much as or more than $20 a month for a single core (in a true virtualization environment like XEN, KVM, VMWARE). You may find some exceptions to this with OpenVZ providers but this is not even full virtualization to begin with and most will limit you to a 'Mhz' instead of a core limit.
TL; DR:
@joe0528 I think the person who needs to change their expectations is you, not the other way around here. It sounds like you are not familiar with the low end at all if you are expecting dedicated resources for the prices you see here, it just isn't going to happen.
my 2 cents.
Cheers!
Translated: "I've not had any problems but reading a 2 year old thread on the internet has made me realise I have all of these problems (which I haven't actually experienced), and now Host X is the worst host ever and I must leave!"
If you need to use 100% CPU all day, get a dedicated server and stop ruining the experience/performance for everyone else.
Depends what full access means. If locking the core, no, you can be sure it is not the case, albeit, you may get around with it as they probably do not monitor the nodes so closely and do not keep such high reserves.
@Jar must know more about this, but I am almost sure you do not get official approval to lock the cores or they have other means of limiting you if this happens.
it is not conceivable a 24 cores dual E5 (with HT) will hold only 24 512 MB droplets as this means about 16 GB ram at most, while such nodes have at least 128 GB ram.
If they keep only 10% idle CPU capacity, you can expect to have at least 5:1 if not 10:1 oversubscribing, which means a CPU core, even half of a core with HT, is NOT dedicated.
In the end, it depends on the policy, we keep 50% idle CPU in most cases of budget servers (12 cores with HT=24):
[root@pm67overzold ~]# uptime
05:53:10 up 12 days, 18:06, 1 user, load average: 10.57, 11.52, 11.38
And over 70% in other cases, IN AVERAGE. This means we have plenty of space for spikes, keep %st close to 0, etc.
In my experience, at least 30% must be kept in reserve, otherwise customers will not be happy due to frequent spikes and sluggishness.
Definitely not dedicated, but you do have full access to the core(s) yes. Much like is standard among hosts on LEB, you're free to use it but everyone using all of it at the same time would create a problem and the reaction to that may vary based on circumstances. Generally speaking, we're quite accommodating about it.
Probability tends to land in your favor on KVM hosts over VZ ones given that no matter the technique, overselling simply cannot be done to the same degree on KVM as on VZ, in my opinion. That said, I still love OpenVZ. I'm definitely not in the "KVM is awesome and OpenVZ sucks" camp.
@joe0528:
Make a virtual machine in Virtual Box that uses all of your CPU cores and tell me how that works.
The root operating system doesn't free up the core and give it to the VM, the VM hypervisor will spawn threads on the host to match up to the VM's CPU requirements so the tasks can be divided up.
For Xen and KVM, each CPU is a qemu or similar process depending on the hypervisor. With that being an application process just like something just say umm nginx, each process gets it's own slice of CPU time. That's how the resources are shared. The hypervisor could spawn the qemu processes on any CPU core so it could cause issues with performance if a neighbor is abusing it as well.
For OpenVZ, all the users processes that are ran in the fancy chroot are just totaled up to make the actual load. Since an app is ran in the jail, it can run on any of the hypervisors cores, so it could get slammed with another users processes or even your own.
If you really want dedicated CPU cores to do your tasks, then you need to use your own server. There is no such thing as dedicated CPU cores.
I am not asking for dedicated CPU. The thing is that RamNode has a requirement about "should not max out 1 full core consistently" even with a 4-cores package. Usually 4-cores package means not maxing out 4 full core consistently in other VPS providers. And it's in the RamNode AUP that I didn't read and I was just wondering why not put this requirement in the package list as a reminder so that I wouldn't bother to go through the ordering process. And of course, they will simply blame me that I didn't read every words in the AUP before ordering.
Here's how it appears to me.
When you get the "4 cores" plan you are paying for 4 cores. If you get a smaller plan, you will only have access to 1 core (or 2)
Now, if you only have access to 1 core, and do something extremely CPU intensive you'll probably max it out, and if you do that for a long time AND cause issues for other customers on that node, you'll get a warning, your VPS will be rebooted, etc. whatever.
If you pay for 4 cores, you have access to 4 cores, and running the same script/task may still use the load of 1 full core, but it will (should?) be spread across all 4 cores so you won't have the "maxing out a full core" issue and therefore it shouldn't be a problem.
If your script only works with 1 core, and it keeps maxing it out, then that's a problem with the script, and even paying for a dedicated server with 4 cores will be a waste of money. If this is the case you should go with something like abusive-cores by drServer
My advice would be to run what you're planning to run on Ramnode, and see what happens. If it doesn't cause issues, then you won't have any problems. If it constantly impacts the node, then Ramnode will deal with it how they see fit (and as defined in the T&Cs)
That's how I see it anyway.
Edit: and if you're wondering why causing the same load on 1 core is different to 25% of the load on each of 4 cores, then (and I'm not 100% sure how containers/virtualisation work) but imagine if 10% of the customers are assigned one core (core A) and the rest are on cores B-J, then if you max out core A, that 10% of customers will suffer.
Honestly it's not worth going down that road. If you do this for everyone who feels that strongly about X policy, the order page becomes a restatement of the policy and then you're in the same boat...people won't read it, it's too long.
For you it's this issue, for someone else it'll be something else and they will feel as strongly as you do about this, or more so. There's someone out there raging about some policy that you consider to be stupid obvious, I guarantee it.
You can't win them all. There is no correct action to please everyone.
That is how the process is right now, only difference is, its not on same page, but u have clickable link to see them before Agree to buy.
Also one rule cant be divided from TOS section and putted before you press Agree because the product you are buying is not for one and single use, it can be used for various things and not all of them require full cpu usage.
Also option is send ticket to sales with any questions you seems are not clear. That will save you more time and money then complaining later.
If you saw their offer in LEB there is always this :
**As always, kindly check their Terms of Service and Acceptable Usage Policy before purchasing any service.**with links.Personally only newbie did not read TOS and AUP section, which is why all newbies learn their lesions after 1st or 2nd vps
@joe0528 You might be able to use all 4 cores on the plan from contabo (price including VAT). It's a German GmbH (=25,000€ minimum capital) so you probably won't have to deal with false promises and whiny comments in a forum with $1 LLC company owners whose mental horizon ends at the borders of their country.
@Lee What's with the disrespect? Go read https://dejure.org/gesetze/BGB/305c.html
Yes, lesson learned. But the point is there will always be newbies. The whole thing is a sophisticated scheme to create the business edge.
Yes, just prove again that if there is a much better deal than usual, there is always a catch. And you always get what you pay for. Pay cheap, get cheap. That's all. I am done with all this.