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.
hosthatch nvme server has more poorly perform than before
I bought a nvme server with 2 cores 8g ram before, when bought it has 600+ score per core, but now it has just 100+, does it stiil have 50% dedicate now?
@hosthatch
https://browser.geekbench.com/v5/cpu/10178042
Comments
Thanks for contacting lowendhelpdesk, I've doubled your ticket number.
Have you opened a ticket to let them know of the performance drop? It sounds like there may be some noisy neighbors on your node or because you purchased this from LET and one of their specials, it is possible the server is just quite loaded -- a lot of VMs need stuck on a server to be able to afford to sell it to you at the price they do. Either way, I am sure if you let them know about your issues they will be happy to help you out. Keep in mind, you may feel this is super important, but when it comes down to it, your VM is still working.. so if you put in a ticket for this, be realistic about your expectations on reply time -- for something like this I would give it a least a week with them.. but if you are patient they will take care of you.
my 2 cents.
Cheers!
Make yabs on cron, every 10minute. Dump cpu core result to a file and after 1 month non-stop yab'ing calculate average result.
Clearly, you must sue.
Definition of insanity, lol..
I wouldn't call the practice insane.
Benchmark junkies buy webhosting just to run yabs non-stop for a month to see a result and, if bad, cancel afterwards.
I have try more than ten times now, all are around 200
thanks ur cpu steal has been doubled.
Yes, it's two days after post the ticket, maybe I can get some reponse after one week, I hope this can let hosthatch know my problem more quick.
Try harder. Give us average from 200 result set. Make 400 set to be on a safe side.
I think it's abusive and stupid to "test" Geekbench and bandwidth so often after buying a VPS if your programs aren't running slow or crashing.
Because you are testing for the sake of the score and not really taking advantage of the VPS.
So, I think you should consider the performance you need, not the results of the so-called tests.
You do not understand.
For some, whether a website or an application runs flawless or not is not the main issue.
For some, it's just raw numbers they want, and OP wants that number. How could he impress his guy friends otherwise?
exactly
if you are not running yabs frequently enough then how provider could know that you need/expect more performance and he should alocate more compute/storage/network resources for you???
What location is this
Yes, I also found this problem
Decided not to renew this year
All performance has dropped twice as compared to the original, and I accidentally tested it yesterday
One less mjj on the node.
"Accidentally tested" my throat.
Yeah, one less mofo on the node. Good for everyone.
This is why you should never buy yearly deals on LET
Yeah, buy 3 year-deals instead. Those are far better.
I also have noticed some slowness, it started in the last 2 months.
EDIT: Nahh please disregard my message, I did some benchmarks and it seems good. Apologies to HostHatch.
Also I want to note that the server that I am in is very reliable, no downtime since last time I formatted it, like 6 months according to runcloud.
Well, depend on the provider & user. Almost 90% of my yearly deals perfectly fine since 2018.
My best guess is that this was throttled because of the constant benchmarks. We very rarely have issues with CPU nowadays. 9/10 of the tickets I have seen in the recent past are users who got throttled to their CPU limit.
The benchmark OP posted also shows the same, about 50% of one of the core's worth of points, which is the plan's dedicated limit.
The extra CPU on top is there for when you need it, and 99% of users have no problems using it when they need it, but "running a benchmark every 30 minutes" is one of those use-cases that probably won't make the cut.
From a provider's perspective, I'd argue the opposite - we see far less abuse on yearly paid servers than on monthly....you can guess why
First, I am very satisfied of your servers, they are really stable and affordable, this problem is just found this days.
But I think your guess is not established. before my first time benchmark these days, my server is always idle as little task runs in, not every xxx minutes benchmark, and they are really stable score 200+ no matter idle for a long time or benchmark every xx minutes, it verified they are not relative.
in addition, my storage plan has the stable score too, But it has 500+ as When bought.
https://browser.geekbench.com/v5/cpu/10186672
The score has too much gap compared with when bought. as a compute plan it is not acceptable now, What's the change?
storage plan score
For the love of the end,
quit running benchmarks. That's all you do, ain't it.
Poor neighbors.
what's your score? Is that nvme plan?
https://browser.geekbench.com/v5/cpu/7191370
The nvme score When bought.