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
First server completely died: confirmed. less than 3 hours after posting here.
So that was a good time to migrate
What are your guesses, will they refund or compensate me or neither, for less than 70% uptime
Yeah, it is clean. Is it just after a boot? You need to capture it right after the drive fails
My guess is that your uptime measurements are incorrect, or something is messed with your BIOS. I was using a basic Mevspace server as a Proxmox hypervisor for a few months and except a few network issues (but still 99%+ up) it was great value for the price. They are the Poland version of Expenzner.
Ryzen 5 4650, 64GB RAM, 2TB NVMe, unmetered 1gbps for 48 Euros/m in 2026 is a deal.
And they provide free IPMI for 4 hours after a few minutes if shit hits the fan. So I have a little
different experience than you, but again, I took the basic server and made sure I have the latest BIOS (had to violate ToS to access it), default settings, no power saving BS, and that the
drive was in good shape. I guess you hit some issue with your hardware combo. Yes, it's not a datacenter grade hardware, but you also don't pay datacenter grade prices
Keep in mind MevSpace is a heaven for stresser operators.
My problem is that:
1) The hardware is dying completely, and unable to boot back up without full server reboot.
2) The SSDs died after 6 days of work on that server.
3) Why does a customer needs to go through all this work, to get a working server they have paid for?
I get that 120 euros is a cheap price for a dedicated server, but also I would at least expect the server to work, not have high end enterprise infrastructure and hardware
they have enterprise nvmes as well, have to pick server type specifically with those though
I used to have their Ryzen 3600 dedicated server. When I bought it, their tech noticed there was a hardware issue, they replaced the server and notified me and I was up and running. After that incident, I've had no problems with their service. It's unfortunate that you are having issues.
If the SSD constantly dying, either it's due for replacement or your workloads is incompatible with the hardware capability.
Pretty common issue on Ryzen systems tbh. Could be a mb/pcie issue or could be NVMe temps and the controller is cooking like an egg and dying, especially if it's under a NIC or something.
Unable to boot back up - as in backup the server (software issue), or the server is frozen (kernel panic) and you have to hard reboot it? Please clarify as it is an important thing to
have or at least troubleshoot the issue. Do you have logs when the server is "dead"?
I'm trying to do my best to help you, and a little bit on the side of Mevspace here, because
I really don't think it's something on their end. I will tell why - here is a simple test.
Run yabs in a loop for 24/7, print the logs to /root/logs.txt . Run it for 24 hours.
Server doctor at your service.
Yeah, temperature matters. I assumed @flexinov checked the nvme temp before posting the thread
NVMe controllers are not cooking machines. I'm not sure what OP is doing, but this must
be over the workload of typical stuff that hardware should do.
"unable to boot back up without full server reboot" - if you mention a boot, that's already a
full server reboot my friend. So please clarify.
@Mevspace I won't be shielding you more than I care, chime in and participate.
I had this issue on standby Proxmox setup, with no VMs, running on top of Debian 13
Temperatures put them in slow cooker status for sure.
Just as a measurement - Can you run yabs on a clean server, Alpine Linux, latest kernel,
Mevspace provides 48h free test of any server. This will make the picture more clear. I'm
100% it's a software issue - but I want to check my feelings against real tests.
I'm also not sure how your measurements were done. Install Proxmox and YOLO?
First you run a 24h benchmark on the disk, CPU, and RAM. Nothing fails? Good.
We have gotten the replacement server, that we are test driving now, after the first one died. After first 2 reboots at the start, server is holding straight. I will update here if anything changes
Run my methodology and I'm sure you will be fine. Ask mods to change the title, this is
just a, how to call it, not the right way to call a provider like that. They are not a scam.
@Mevspace where are my free credits? /s
Typical MJJ behavior!
I never called them a scam lol? Just the worst server provider in my opinion. There is plenty of people who share my frustration in the same thread
Again not true, hetzner have drop in IPMI to none server boards, they have full documentary on youtube about this, when LTT visits to build a server.
Rule of thumb, expect any low-end priced server to have none enterprise hardware.
None of my Hetzner servers, i9, Ryzen, have IPMI. And I don't think I've ever had a Hetzner server with IPMI in the past. They have KVM console that you can request that they will temporarily attach to your server. Maybe you're thinking about their remote reboot feature that was shown in the LTT video. Or maybe I'm misunderstanding your sentence and you mean none of them have drop in IPMI.
I could be mistaken but i've had IPMI with them so maybe that's caused me to re-write the truth, my bad could be mistaken
Hi, MEVSPACE here on a Sunday evening/night.
First of all, I'm sorry to hear that you've had problems with our services. Please send me a DM with your account or ticket details and I will personally track down what happened in your case. If I'm not able to resolve it directly, we will treat it as a post-mortem case and investigate what went wrong and what we can improve internally.
I'm not going to argue here whether this was a software or hardware problem, because based on the information available it simply requires a deeper investigation.
Regarding our hardware: we operate thousands of physical servers, and we do our best to identify and eliminate faulty components before they affect customers. At this scale, however, hardware failures do happen. What matters to us is how we react when they do.
We know that sometimes the fault is on our side, and we don't hide from that. We try to learn from every recurring or unusual failure pattern and improve our testing and operational procedures accordingly.
A good historical example was one of our Ryzen 3600 batches. We received customer reports about instability, while our standard hardware tests were unable to reproduce the problem. After a deeper investigation, we eventually identified a very specific workload/instruction pattern that could trigger CPU instability. It turned out that more than 3.5% of CPUs from that particular batch were affected. That experience resulted in changes to the way we diagnose similar cases today.
This is exactly why individual reports are important to us. Sometimes what initially looks like an isolated customer issue can reveal a wider hardware or operational problem.
Regarding the components we use: yes, Goodram memory is used in a large part of our server fleet. Our standard configurations are designed around a balance between performance, reliability and the pricing level for which MEVSPACE is known. We know hardware can fail. When it does, we need to find the faulty part and replace it as quickly as possible.
For customers who require specific enterprise-grade components, particular vendors, ECC memory, storage models or other specialised configurations, we can also build such solutions, although naturally they come at a different price point.
We know we are not perfect. What I can promise is that we take cases like this seriously, investigate them and use what we learn to improve the service.
Please DM me the details and I’ll take a look. And after that, you can start hating me again
adam
They have them on the branded lines like DX-line
i do think have a same when i first server for my friend we find ram issues and they had to all server since that time which was in first 5 days it was pretty good for them to replace all of them in couple of hours, for such a price i don't mind that, anyway i had my tool to backup everything for multiple gateway so i think it's all amount collaboration with provider and have some patience
from my own experience (using them to configure some for my friend) they are good, they even start investige all servers when one of his server was experience ram issues
i'm using mevspace since 2019, no issue at all,
they are good, and professional, they replace the burnt ram, and hdd, you must setup raid 1 always, their support is now faster as compare to last year,
in that competitive pricing, they must need to improve the network connectivity slightly. overall 9/10 score