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
JP location is not available for selection.
Thanks
Order number: 2085376485
Hello guys!
Could you please provide an estimate for the migration from de11 to fra2, and when the server is expected to be operational? I need this information to decide whether to set up my service elsewhere or wait.
@rarecloud We all know that all providers are overselling VPS. But an E5 CPU got only 350 points in GB5 test, that's insane and really RARE I've nerver seen. Since my order has been canceled for over 20 days, my refund has still not been processed yet. I feel like this is a scam. Please do something. Invoice #V4V-9028
@rarecloud
Hi, could you please expedite the ticket for me? It has been two weeks already
VZH-193602
And what makes you think it's just your ticket they are ignoring?
Yes, I’ve noticed that many others have reported the same issue. But for now, there’s not much I can do besides keep following up
Order Number: 9178977383
After migration only ipv6 for 2 of 3 servers. ASN changed to a worst one. No support at all. Just ignoring tickets for days.
@rarecloud
after migration to Frankfurt, my service stopped working as before, there may be some issues with routing / with new ip address / with something else, I dont know. Could you please move me to another location, such as RO - EU. ticket BYZ-051408
Hello everyone.
Finally, on December 22nd (meaning there was a three-week downtime), my VDS migrated to Frankfurt with a new IP address. Surprisingly, my data was not lost. Thank you, I'm waiting for the end of my paid period to say goodbye.
Order Number: 6108565638
@rarecloud
Before migrating from Düsseldorf to Frankfurt, I had the following scores on Geekbench 6:
Single Core 1621
Multi Core 2834
Yabs: https://lowendtalk.com/discussion/comment/4526324#Comment_4526324
Now I see a ‘QEMU Virtual CPU version 2.5+’ with 2 @ 2895.650 MHz
Single Core 871
Multi Core 1576
Yabs:
I cannot confirm the ‘major performance improvement’ described in the email. In fact, it is a performance loss of almost 50%.
I have the "Plus" with 2 vCPU / 4GB RAM / 65GB NVMe
how long is my windows vps get online?
Official Update: Germany and Netherlands Migration Completed
We want to share a clear and honest update regarding Germany and the Netherlands.
Migration Status
The migrations are now fully completed:
Acknowledging Past Issues
We are fully aware that in the past we had serious issues in these locations.
In the Netherlands, storage failures led to data loss.
In Germany, there were recurring network problems.
These issues came from old, problematic datacenters and low-end hardware choices and yes, they damaged our reputation.
The most important part is this: we acknowledge those problems, we take responsibility for them, and we acted.
What Changed
We have completely retired the old datacenters with known issues and migrated all affected clients to brand-new infrastructure.
Every customer, including low-end plans priced extremely cheaply, has effectively received a major hardware and datacenter upgrade.
We are now running on modern, high-end servers and significantly more professional datacenter environments. This was not done for marketing or appearances, but to permanently fix the root causes of the problems we had.
About Our Forum Activity
We also want to address something openly.
Over the past weeks, due to the scale of the problems and the workload involved, we were not as responsive on the forum as we should have been. That part is on us.
We made a conscious decision to focus on the technical side first: stabilizing infrastructure, migrating data, fixing servers, and helping our team resolve issues as fast as possible. As a result, some forum comments were answered late or not answered at all.
We understand this is not ideal and we apologize for it. At the same time, we ask for a bit of understanding.
Who We Are and How We Work
We are still a small company and a small team, yet we currently serve 3000+ customers across 12 datacenters worldwide.
Many of us wear multiple hats. The same people you talk to on the forum are also involved in management, technical decision-making, server deployments, migrations, and daily operations.
We all do more than one job here, and we do it because we genuinely want to offer the best possible services at the best possible prices.
There is always a trade-off. You cannot realistically expect premium, brand-new hardware, professional datacenters, and ultra-low prices without some compromises.
What we can promise is that when things go wrong, we do not hide, we do not disappear, and we do not give up.
Black Friday and the Increased Load
We would also like to address the Black Friday campaign.
The campaign performed far better than we expected. While this was great to see, it also led to more orders than anticipated, a significantly higher support load, and all of this overlapping with large-scale migrations.
From the outside, this may have looked like chaos. Internally, however, we always had a clear plan and focused on fixing things properly rather than rushing incomplete solutions.
If there are still tickets waiting for a response, they will be answered.
We have expanded our support team for the third time in the last 30 days, because our goal is to deliver reliable services, not excuses.
Looking Forward
We are not perfect. We made mistakes.
But we learn, we improve, and we keep moving forward.
Our goal remains very clear: to grow into a serious, reliable provider that continuously improves and delivers solid services, even at prices that are, frankly, far below what this level of hardware and infrastructure usually costs.
Thank you to everyone who stayed with us, challenged us, and pushed us to do better.
We are here for the long run.
P.S. Tomorrow we will draw and publish the results of the Black Friday raffle.
Hi @berndy2001, thanks for the detailed benchmarks and for taking the time to post them.
Sorry about this. What you’re seeing is a known issue related to the latest Ubuntu 24 updates on the nodes, combined with how CPU mode is currently handled in Virtualizor.
Until recently, when CPU Mode was set to “default”, KVM would correctly expose the host CPU model. After recent changes, this is no longer consistent, and some VPSes end up showing QEMU Virtual CPU instead of the real host CPU, which also impacts performance (including AES-NI).
To fix this, the CPU model needs to be explicitly set to Host Passthrough at the VPS level. I’m already discussing this with our CTO to implement a permanent, node-level solution. Until then, please open a support ticket and ask our team to switch the CPU model to Host Passthrough for your VPS, and this will be resolved.
Additionally, I’d like to ask for a bit of patience. We have just finished migrating a large number of VMs to brand-new hosts. During this period:
• there is higher than usual load and IO due to the migration traffic itself;
• many customers run benchmarks at the same time right after landing on the new nodes.
This kind of “benchmark storm” does not reflect normal day-to-day usage and can temporarily skew results, making the servers appear more affected than they actually are under real workloads.
That said, your point is valid, and the CPU model issue should not happen. Once Host Passthrough is enabled and the initial migration load settles, you should see performance in line with (or better than) what was advertised.
If anything still looks off after that, please let us know and we’ll investigate further.
🎄 Merry Christmas, LET! 🎅
We're back with some absolutely insane flash sales that you won't find anywhere else. These prices are so low, you might think we've lost our minds, but we're celebrating Christmas by giving you the best deals possible!
🔥 FLASH SALES - Limited Stock, Maximum Value!
💥 Advanced VPS: €12.50/Year (75% OFF)!
2.5GB RAM + 5TB Bandwidth for just €12.50/year. This is one of our MOST POPULAR plans at an absolutely INSANE price!
Specs:
- 1 vCPU (High Performance)
- 2.5GB RAM ⚡
- 45GB NVMe Storage
- 5TB/month Bandwidth 🚀
- 1x IPv4 & /64 IPv6 Subnet
- Locations: EU/US/JP/HK (including our brand-new Hong Kong location!)
Coupon:
BF25FLASHADV12⚠️ Only 10 uses available!
➡️ Order Advanced VPS Here
⚡ Plus VPS: €20/Year (79% OFF)!
4GB RAM + 5TB Bandwidth for just €20/year. This is 79% OFF regular pricing!
Specs:
- 2 vCPU (High Performance)
- 4GB RAM
- 65GB NVMe Storage
- 5TB/month Bandwidth
- 1x IPv4 & /64 IPv6 Subnet
- Locations: EU/US/JP/HK
Coupon:
BF25FLASHPLUS20⚠️ Only 30 uses available!
➡️ Order Plus VPS Here
🔥 Pro VPS: €30/Year (85% OFF)!
8GB RAM + 7TB Bandwidth for just €30/year. This is 85% OFF! The biggest discount in this thread!
Specs:
- 4 vCPU (High Performance)
- 8GB RAM
- 100GB NVMe Storage
- 7TB/month Bandwidth
- 1x IPv4 & /64 IPv6 Subnet
- Locations: EU/US/JP/HK
Coupon:
BF25FLASHPRO30⚠️ Only 30 uses available!
➡️ Order Pro VPS Here
🚀 Elite VPS: €59/Year (83% OFF)!
16GB RAM + 10TB Bandwidth for just €59/year. Our BIGGEST flash sale ever! Maximum enterprise power at an impossible price.
Specs:
- 4 vCPU (High Performance)
- 16GB RAM
- 150GB NVMe Storage
- 10TB/month Bandwidth
- 1x IPv4 & /64 IPv6 Subnet
- Locations: EU/US/JP/HK
Coupon:
BF25FLASHELITE59⚠️ Only 10 uses available!
➡️ Order Elite VPS Here
🌍 Why These Deals Are Even More Insane Than They Look
We just launched three brand-new premium locations that make these flash sales even more valuable: 🇭🇰 Hong Kong, 🇩🇪 Frankfurt, Germany, and 🇳🇱 The Hague, Netherlands. All three are powered by M247 datacenters with AMD EPYC Genoa 9254 processors, brand-new servers, RAID10 NVMe SSD storage, and massive bandwidth included.
Here's the thing: You won't find prices this low for hardware this good anywhere else. For example, take our Advanced VPS flash sale at €12.50/year. That's 2.5GB RAM + 5TB bandwidth for less than €1.05/month. I don't think there's a single provider out there offering more than 1GB RAM and 500GB-1TB bandwidth in Hong Kong for €10-15/year. We're giving you 2.5GB RAM + 5TB bandwidth at the same price point, in premium datacenters, on brand-new AMD EPYC Genoa 9254 servers with RAID10 NVMe storage.
That's the value we're offering. Premium hardware, massive resources, brand-new infrastructure, all at prices that seem impossible.
🎄 Plus: Our Christmas Sale Continues!
On top of these flash sales, we're continuing our incredible Black Friday deals for Christmas: the same massive discounts that thousands of people loved and ordered.
All available in our new premium locations (Hong Kong, Frankfurt, The Hague) plus all our existing locations: Tokyo, London, Silicon Valley, Phoenix, Dallas, Bucharest, Timișoara, and Constanța.
See all our Christmas deals in one place: https://rarecloud.io/active-promo/
Merry Christmas, everyone! 🎄🎅
These flash sales won't last long, so grab them while you can!
🎁 Want Even More Value? Quote & Comment for Extra Benefits!
We want to hear from you! If you order any of our deals (flash sales or regular Christmas offers), quote this post and comment with your order number, and we'll add extra value to your service:
Flash Sale Orders: Get +1 month FREE added to your next due date! You pay for 12 months, get 13 months of service. That's an extra month of our best deals at no extra cost!
Regular Christmas Sale Orders: Get +1GB RAM or +1TB bandwidth added to your service permanently! More resources, same incredible price.
How it works:
1. Order any of our deals (flash sales or regular Christmas offers)
2. Quote this post and comment with your order number
3. We'll verify and add your bonus within 24 hours
This is our way of saying thank you for being part of the LET community and for choosing RareCloud. Happy holidays! 🎄
Do you have any issues? Can you explain more? Can you DM me or even better, open a support ticket so we can check it?
Thanks for sharing your experience. We understand the frustration, and we won’t try to deny or sugarcoat what happened.
You are right: in the past, especially in Germany and the Netherlands, we had serious issues. There were network problems, storage failures, long downtimes, and in some cases even data loss. Those problems came from the fact that RareCloud started as a small hosting company, built on low-end infrastructure, with a lot of enthusiasm but not enough maturity in hardware choices and datacenter partners.
That part is on us.
What matters now is what we did about it.
Over the last months, we made the decision to completely abandon the old infrastructure that was causing these problems. We retired problematic datacenters, replaced low-end servers, and migrated everything to new locations with modern hardware and much stricter standards. The recent migrations to Frankfurt and The Hague were done specifically to fix these issues at their root, not to temporarily patch them.
We also understand why an IP change can feel like “losing the server” for certain use cases. Unfortunately, keeping the old IPs was not possible once the old datacenters were retired. The alternative would have been to keep broken infrastructure running, which would have only resulted in more downtime in the future.
Regarding responsiveness, you are also right to be upset. During this period, our focus was heavily on the technical side: restoring access, migrating data, rebuilding servers, and coordinating with datacenters. That meant we were slower on the forum and in some tickets than we should have been.
This is not an excuse, but it is the reality of a small team handling large-scale issues at the same time.
Now, regarding the RAID comment.
P.S. About “the lack of RAID on the boot drive”: this is a common misunderstanding, especially if you don’t work directly with physical hardware. On modern NVMe-based servers, traditional hardware RAID controllers are not used. NVMe drives are designed to be managed either as single drives or via software RAID. Using classic hardware RAID with NVMe is not standard practice and, in many cases, not even possible. Most modern providers rely on software-level redundancy, backups, and replication, rather than hardware RAID on NVMe boot devices.
We are not claiming perfection. We made mistakes, and some customers understandably decided to move on. All we can say is that we learned the hard way, invested heavily, and rebuilt our infrastructure specifically to avoid repeating the same problems.
We appreciate you acknowledging that your data was not lost during the migration, and we genuinely wish you the best going forward, whether you stay with us or choose another provider.
Regarding compensation
We mentioned by email during the migrations that we are offering between 2 weeks and up to 2 months of extra service as compensation for downtime, depending on the situation.
However, considering the number of issues and the long downtime you personally experienced, we would like to go further.
We would like to extend your next due date by an additional 1 full year, free of charge.
This is not about money for us. We would much rather give you the chance to test the new infrastructure for a full year, and hopefully allow us to prove, to you and to others reading this forum, that we have genuinely evolved.
Please accept our apologies and this compensation.
Our goal is not to convince with words, but to let the service itself show that the effort, sacrifices, and changes were worth it in the end.
Please reply here or send us a DM with any detail that helps us identify your service, such as your order number, the current IP address of your VPS, or the email address associated with your account.
I've checked the ticket and it seems that my colleagues already solved your issue.
Can you please share the ticket IDs so we can check them?
Official Update: Germany and Netherlands Migration Completed
We want to share a clear and honest update regarding Germany and the Netherlands.
Migration Status
The migrations are now fully completed:
Acknowledging Past Issues
We are fully aware that in the past we had serious issues in these locations.
In the Netherlands, storage failures led to data loss.
In Germany, there were recurring network problems.
These issues came from old, problematic datacenters and low-end hardware choices and yes, they damaged our reputation.
The most important part is this: we acknowledge those problems, we take responsibility for them, and we acted.
What Changed
We have completely retired the old datacenters with known issues and migrated all affected clients to brand-new infrastructure.
Every customer, including low-end plans priced extremely cheaply, has effectively received a major hardware and datacenter upgrade.
We are now running on modern, high-end servers and significantly more professional datacenter environments. This was not done for marketing or appearances, but to permanently fix the root causes of the problems we had.
About Our Forum Activity
We also want to address something openly.
Over the past weeks, due to the scale of the problems and the workload involved, we were not as responsive on the forum as we should have been. That part is on us.
We made a conscious decision to focus on the technical side first: stabilizing infrastructure, migrating data, fixing servers, and helping our team resolve issues as fast as possible. As a result, some forum comments were answered late or not answered at all.
We understand this is not ideal and we apologize for it. At the same time, we ask for a bit of understanding.
Who We Are and How We Work
We are still a small company and a small team, yet we currently serve 3000+ customers across 12 datacenters worldwide.
Many of us wear multiple hats. The same people you talk to on the forum are also involved in management, technical decision-making, server deployments, migrations, and daily operations.
We all do more than one job here, and we do it because we genuinely want to offer the best possible services at the best possible prices.
There is always a trade-off. You cannot realistically expect premium, brand-new hardware, professional datacenters, and ultra-low prices without some compromises.
What we can promise is that when things go wrong, we do not hide, we do not disappear, and we do not give up.
Black Friday and the Increased Load
We would also like to address the Black Friday campaign.
The campaign performed far better than we expected. While this was great to see, it also led to more orders than anticipated, a significantly higher support load, and all of this overlapping with large-scale migrations.
From the outside, this may have looked like chaos. Internally, however, we always had a clear plan and focused on fixing things properly rather than rushing incomplete solutions.
If there are still tickets waiting for a response, they will be answered.
We have expanded our support team for the third time in the last 30 days, because our goal is to deliver reliable services, not excuses.
Looking Forward
We are not perfect. We made mistakes.
But we learn, we improve, and we keep moving forward.
Our goal remains very clear: to grow into a serious, reliable provider that continuously improves and delivers solid services, even at prices that are, frankly, far below what this level of hardware and infrastructure usually costs.
Thank you to everyone who stayed with us, challenged us, and pushed us to do better.
We are here for the long run.
P.S. Tomorrow we will draw and publish the results of the Black Friday raffle.
Have you opened a support ticket asking for refund?
please give me the ticket ID and I will check with my colleagues.
Thanks
HK test ip
85.121.244.2
here's the ticket ID: #LXR-875771. Merry Christmas.
@rarecloud, put some stock in Japan please?
Still waiting for the new servers to arrive there... sorry
Ah, so probably 2026...will wait.
Invoice #V4V-9737