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
Maybe, but the point was whether MXlinux is an exotic distro almost nobody uses or even knows about. Most page views ~ Many know about and it's definitely not an exotic niche distro.
Plus anyway that whole "which distro" point was but a nonsensical point by someone embedded within another attack attempt. Of bloody course I wouldn't use MXlinux on a purchased real (not local) VPS but Devuan or alpine.
@dev_vps and others
UPDATE
I now have got confirmation from the person in charge of virtualization at Contabo who was responsible for creating my test VMs that those VMs were normal VMs, just like those sold to customers, with exactly the same specs and limits, no difference whatsoever except for the fact that the nodes with my test VMs were (almost) empty.
Can that statement be trusted? I think yes, among other reasons for the fact that Contabo regards my benchmarks as basically for internal use (the fact that I publish them on LET is allowed but seems to not be their goal).
At the end every one must decide for themselves.
Update
just got confirmation that the girth of my e-dick far exceeds the common average.
Can this piece of information be trusted?
Yes it can.
Did you finish the test for @dev_vps ? I think you could share the result to backing up your & your friend statement..
waiting too
You are really funny. First you thank the posts of the "jsg's benchmark is sh_tty, worthless, and crap" gang ... and then you ask me for the results of a current benchmark.
But, tough luck. Before I could say 'yes' you'd first need to ask the user for whom I do this benchmark. If and only if he OK's me publishing the results, you can ask me.
FYI: I'll be done probably tomorrow morning (central Europe tz).
Btw, me, the oh so evil guy nevar listening to, let alone thinking about, constructive criticism (as if there was any ...) and nevar evar admitting being wrong in no matter what (hint: actually I admitted an error twice and even apologized) am ... tada, surprise! ... thinking about extending - or as some would say "enhancing" - my benchmark program ("thinking" as in "concretely planning and already laying out design details). Maybe I'll even make it publicly available, but currently I'm not feeling like it; one shouldn't reward nasty a__holes.
@jsg
Thank you for your help with benchmark tests for vps.
Kindly share the benchmark test results here. This will help to compare the benchmark tests.
Thank you @dev_vps for allowing the result to be shared. And thank to @jsg for providing the benchmark.
@jsg
I appreciate your help with benchmark tests.
You had the VPS for over 36 hours. The access password has been reset.
Kindly share the benchmark results you have collected so far.
Thanks again.
Best regards
PS: The current Central Europe TZ time is 3pm
The .bash_history file may (or may not) also make for interesting reading.
I needed some more time because I saw considerable spread and wanted to have more runs for you. And now you treat me like someone who wants to extend a free ride for his own advantage.
I never, no exceptions, use a test VM for anything but testing/benchmarking.
My standard minimum test period is 48 hours for a reason. But I'll admit that I first tried to go faster, which failed, so I had to ask for more time. My bad.
Now I'll compile the results I managed to collect and publish them a bit later.
It's always the small ones that feel the need to prove something...
I really appreciate your help with the benchmark tests. And for that reason, I will let any negatives you mentioned pass.
I am sorry if you feel that way. But these are your words, not mine.
I am a data scientist by profession, and in my field numbers speak for themselves. Based on my limited testing, the disk I/O is limited for small file sizes. For example, compare disk i/o for small block sizes with that of Host Hatch.
Thank you, again.
Here are the results:
4 KB
256 KB
4 MB
Notes:
The node appeared to be very busy (and possibly crowded).
I mainly focused on smaller sized probes (4K and 256K) because those are typically more significant for a DB or dynamic (DB driven) website server.
While the system did meet what my guesstimate was (600+ MB/s in direct/sync mode) I'm somewhat disappointed, especially by the very large spread with small probe size.
I don't know the reasons, maybe an overcrowded node, maybe abusers on the node, maybe quite unbalanced occupation, but I would strongly suggest (and highly likely will) Contabo to get their node management and statistics improved.
Due to sudden early test termination some result sets are smaller than desirable.
Summary: Not a crappy server at all but certainly not what could and should be expected from the node hardware. The NVMe hardware is great but it's sad to see what negative factors (or node management) make of that great hardware in terms of effective performance on the VPS level. Note though that I've seen much better nodes (also in production). But still, there should be no subpar nodes with a top-provider.
Advice: N/A
Further private tests or benchmarks: N/A
gets some good bench results:
OH YEAH SOLDDDDD THIS PROVIDER GOOD SO YUM , WOW THEY ARE 100% KOSHER VPS
got some bad results:
hmmmmm.
hmmmmmmmmmm.
lets look at this from an anal-ysis. hmm it appears to be bad. it appears to be....it appears to be.....but inconclusive. inconclusive ok? why no enough time? why?
hysterical-laughter.gif
Having customer VPS on crowded (busy) node
Test VPS on empty node
enough said.
Thank you @jsg for your unbiased analysis from benchmark testing of a customer VPS.
Respect. ✊
PS : I am thinking of returning of the VPS (14 day money back guarantee). There are other providers who are better suited for my use cases (coding and testing of machine learning algorithms)
Not really. For one: show me the benchmarker who say "no" when being offered a pre-launch VPS/VDS on an almost empty node.
Plus I did (and still do) quite a few benchmarks on other Contabo VPS (which I purchased) and I never saw that level of difference.
Plus it's actually useful to sometimes do pre-launch tests because it offers some (rarely available) insights by being able to compare (almost) empty node performance vs. production performance which allows one to derive quite some interesting and valuable information (and to share that with the community).
And btw, if I approach Contabo with solid data under my belt my chances to convince them to fine tune some screws are probably much higher than someone approaching them with just some complaining and little relevant - and comparative - data.
And at the end of the day I do my reviews for our community - and that's where my loyalty is - and not for this or that provider.
So? I don't care. I'm not doing my benchmarks to drive people to buy from a particular provider, nor btw. to drive people away. That I'm fair and as unbiased as is humanly possible, is proven. I made a very positive review (because his VPS I benchmarked was really, really great) for a provider whom, as almost everybody here knows, I detest.
You buy or buy not or keep or keep not a VPS? Me not care, just do what's best for you.
Totally agree with you that running tests on a VPS (on empty node) is pretty useful. But only for the provider.
From customer point of view, it has little to no value, as it presents the best case scenario to the customer.
It is like driving a car on an empty highway vs crowded highway.
You have my respect and sincere appreciation for this.
We have a "Thanks" links at LET posts for saying "thank you". Btw. you have thanked me way more than often enough.
The 4KB numbers are extremely high yet being called poor and overcrowded. I know those numbers are bogus, but even if they were real, 3GB/s at 4KB is hardcore good. If you've "seen much better nodes" than a 2800MB/s (up to 7700MB/s!!!) at 4KB, we'd all like to know.
Hi everyone! The reason I'm posting and have joined the website is that I've recently bought a Contabo VPS NVMe and the disk I/O speeds reported with standard linux tools is ATROCIOUS. I've yet to run a test with nench though. So I stumbled on the website after googling my problem. And unfortunately my fears have been confirmed - Contabo is definitely trying to rip their customers off and really the statements about "super fast NVMes" are just unbelievable. This feels like a scam.
I went with Contabo because they're so cheap and the hardware specs looked impressive so I thought this was a great deal. But yeah, if it sounds too good to be true then it probably is as they say. They're still ridiculously cheap though. The price is maybe 50% cheaper than the other budget options.
I'm also very sceptical of their Trustpilot rating. Compared to their competitors, such as OVH or Digitalocean, they have a stellar rating (I thought Digitalocean was actually supposed to be a bit more decent, but they're rated 2.2 out of 5 stars). But when you look at who is actually posting the positive reviews (and there are a lot of them), it's users who have most of the time not posted any other reviews AND they were redirected. Now, why would so many people post 5 star reviews only for this particular business out of their good will? Who would bother? I know I wouldn't, even if I really enjoyed some service or website. I think there must have been some incentive, a discount? Some sort of shady stuff like that.
Which brings me to my last point which is that I've found this whole back-and-forth between jsg and others really annoying, or rather jsg's responses. What sort of self-styled "Server review king" does one have to be to keep BSing for 6 pages, burst out with indignation at the mere suggestion one may be wrong and keep disputing everyone else's experience?
I think there are only two explanations, either jsg has received some sort of financial "reward" from Contabo OR he's so full of himself (certainly takes a lot of modesty to call yourself a "King") that he can't admit he's been tricked by the quite obviously dishonest company (and I'm sure they're not the only ones in the low-end server industry). Thanks for the benchmarks guys! I was really confused about what was going on.
Seriously... A "grain of salt" - the "contact" is "pretty high up". Can't make this up.
https://www.lowendtalk.com/discussion/173716/jsg-the-server-review-king-can-you-trust-him-is-yabs-misleading-you#latest
Oh, sorry, so he was actually given the title. Thanks for the link! Yeah, he's certainly being very defensive and something's not quite right. At least when it comes to Contabo, you can add my experience to the list of dissatisfied customers who feel a bit like they've been scammed. Certainly not a company to be raving about, which is what @jsg has been doing in this thread. (myself, I certainly didn't have high expectations to begin with, given the price bracket)
As usual - if you want good german provider then go with netcup, php-friends or hetzner. They wont disappoint.
Thanks for the recommendations! I was actually specifically looking for the most budget option, haha (I wanted a cheap set-up to play around with). But well, you get what you pay for (or less than that)...I've had a look at netcup, I like that they're using Ryzen GPUs in their more expensive package.
he's a server review king. where are your creds?
Well I dont know what specs you need, but also check Scaleway Stardust. Its 0.37euro/mo if you remove flexible ip/ipv4.
1vCore AMD EPYC
1GB ram
200Mbps-1Gbps network
10GB NVMe
Amsterdam location
Good enough for basic game server, web server...
Interesting, thanks! I actually need at least 6GB RAM as I need to run a desktop environment.