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.
How much percentage of one core is given in VM usually??
titaniumboy
Member
in General
Hi, I'm wondering how much percentage of each core is given in VM usually. For example if a VM has 4 cores, than usually how much percentage will be given to VM? Is it 50% or lesser? And if 100% is given then it's considered a dedicated CPU VM?
Comments
It depends from one provider to another.
At HAZI.ro, for VPS we give 100% for each core but if sustained usage is above 50% we might have a discussion.
For VDS, you can use 100% 24/7 without any issue. (well, not really - no one want you to mine crypto or something similar on his infra).
users are given vCore and not the actual full core.
it depends on provider. 1vCore = 1 thread of 1 core or even less.
420%
depends from provider to provider.
I recall some years ago Hostdoc sold a VPS with more 1 more vCore than his server had hardware threads.
he was shamed and hence deadpooled with embarrassment.
today we have c1vhosting advertising "2 vCores" but due to their "software AI" causing 50.0 ST on a heavy workload, it essentially means one can only have access to the power of 1 hardware thread at any moment. it is claimed to be "dynamic allocation" but so far it appears to me more static than dynamic.
Literally Hostdoc deadpooled just because they sold a VPS with 1 more vCore than his hardware threads? What's the issue with that? I have heard many providers oversell their CPU. So why is that an issue?
7%
A proper deadpool.
don't ask me ask Hostdoc
different provider may different definition,better run geekbench and compare with average same cpu score
Depends on provider and what level the node is oversold.
I've been testing kvm and 1 vcpu is basically 1 thread / cpu core with the default values. So if my observation is correct, unless specified as dedicated, 1 vcpu will utilize as much as 100% of a thread / cpu corebas long as the server has resources to handle it. In a shared environment this can therefore be less than 100%. Just to be clear, the reason I used 100% thread / cpu core is because some cores if I understand correctly can sometimes handle more than 1 thread thus to be clear I had to add that in. If have somehow made a flawed dedication based on my testing, I always appreciate standing corrected by someone who understands architecture more than I do. But from my current understanding most providers here are doing alot of overselling to keep the pricing down (not bad just amazing when I think about it)
Naaa, with today’s test methods, usually customers will sniff you if you oversell cpu, we stay away from that, and from the tests we say, most do the same.
I personally think that this is disrespectful forwards your customer. Better to have a discution on with him if he needs high cpu load 100% of the time, and craft an offer that he can actually use, rather then oversell and get caught, then the drama kicks in on the forum.
I really like those who approach us and ask, hey I need X cores at 100% all time, can we do this on your network ?
But, anyone is free to do as it pleases.
Most providers permit 25 to 35% average CPU core utilization over 24 hour period.
Few allow around 45-50% utilization
Very few allow unmetered utilization but CPU resource allocation is dynamically adjusted (in other words rationed -- hint --> multicore GB6 is poor for such providers).
Only one provider, I have come cross, for a VPS, he said you can use CPU resources up to 75% average.
VDS is unrestricted usage subject to non allowed tasks such as block chain, crypto etc
if I have to pick one of the best deals, CPU resource wise, I have seen here on LET, is 4 GB memory, 1 core dedicated ryzen for $3.75 per month
Maybe I made a bad assumption here in terms of overselling, perhaps it's idling then? I could be assuming that providers are using systems with less than 32 cores a lot of the time and so the price even when I look at cheap hosts like OVH doesn't add up as to how they could be selling at the cheap prices they sell. It's a mystery I haven't been able to solve in the years being here - I suppose it keeps life exciting now thinking about it
P.S I spend hundreds a year on idling servers but I'm working on rehabilitation
If im renting a vds and im using 100% of the cores 24/7 the hoster shouldn't care at all unless he gets a court order or something similar
Maybe he means artificial Italian when hes offering AI cores?
At yoursunny summer host, you can use unlimited CPU cores in our Antarctica IPv9 VPS.
While each VPS comes with 1~4 cores, the computational power of each core is essentially unlimited.
Your compute tasks are distributed among all the water molecules in Antarctica icebergs, which is technically finite but practically unlimited.
AI powered?
You have hired some penguins as technicians, I hear they are pretty good at administering Linux?