NEWS FLASH New Rack Nerd Deals on KVM VPS

[kenn10]

RackNerd | KVM VPS Special Deals

www.racknerd.com

 

[krzykat]

Has anyone ever used this business model of services in production or even a backup PBX successfully? I've been burned in the past, and kind of err on the side of if its too good to be true, then it is. If I can't see how the business model actually turns a profit, then I think its best to be warry.

 

[hecatae]

Nothing wrong with my Texas or Toronto servers, I would recommend going via https://lowendtalk.com/discussion/206701/new-location-now-live-toronto-canada/p1 and commenting your order to double your bandwidth instead of the link above.

 

[kenn10]

I use the Ashburn servers with no issue. They have been upgrading their servers pretty regularly as well. I have FusionPBX running on a 2 GB server and have had no issues. I did get a 4 GB server at Colo Crossing but it has been crap, even after they moved me to a different node. The computer part was OK but the actual internet connection to it is terrible with dropped packets and jitter.

In the early days of Rack Nerds, I had a server crash completely and had to rebuild but I have had no trouble with them in the past 18 months. Things seem to be well managed now.

 

[wardmundy]

I've had much the same experience as @kenn10 with both RackNerd and ColoCrossing. We have five running at RackNerd and have never had a problem.

 

[kameleon]

I currently have 8 RackNerd vms across the US. Zero issues with any of them, or at least no issues that would be RN caused, I have caused many issues lol

 

[kenn10]

I snagged a 3-cpu, 3-1/2 Gb VM out of Ashland (in VA outside DC) and took the memory up to 5 GB. Performance is impressive. I'll be building a multi-domain FusionPBX on this system.

 

[hecatae]

https://lowendtalk.com/discussion/213430/dedirock-storage-wars-returns-unreal-pricing#latest

2 TB space, what a great deal for a recordings server

 

[phinphan]

I have a simple cheap less than $10/year server with dedirock. I use it for pangolin reverse proxy. I am very happy with it.

 

[Dempan]

Has anyone ever used this business model of services in production or even a backup PBX successfully? I've been burned in the past, and kind of err on the side of if its too good to be true, then it is. If I can't see how the business model actually turns a profit, then I think its best to be warry.

Yeah, I’ve run into the same thing before and learned the hard way as well.

If a PBX or “managed services” model looks too cheap or too all-in-one, there’s usually a reason for it. Either they’re heavily subsidising acquisition costs, cutting corners on support, or banking on churn not catching up with them too fast.

What I tend to look for now is how they handle scale and failure points. Like, what happens when something breaks at 2am, or when you need to port numbers out, or when they suddenly change pricing. That’s usually where the real business model shows itself.

Also agree with you on transparency—if I can’t roughly map out how they’re making money per customer, I get cautious pretty quickly. Doesn’t mean it’s always bad, but it definitely needs extra scrutiny before trusting anything production-critical to it.

On a related note, I’ve found that integrating tools that streamline critical workflows can save a lot of headaches down the line. For example, using integrate docusign makes contract handling seamless and reliable, so you’re not scrambling when signatures or approvals are time-sensitive. It’s one of those things where investing a little upfront in trusted integrations pays off massively when you scale.

 

[kenn10]

I suspect a lot of these VM providers are using old equipment racked at the data center. As newer stuff debuts, it is priced for higher cost services.

I've had excellent luck with Rack Nerd and use them for both production and backup production servers. I'll admit there was a system failure a few years ago but they moved me to another location in the same data center and I was back up pretty quickly. They don't offer backups or snapshots so you need to have good backup files kept locally.

 

[tycho]

I've had a very modest RackNerd server since I read about it in this forum in November '21. I've updated it periodically, moving it to newer IPBX iterations over time, but it has -- for me -- been rock solid 24 x 7. It's production, but not mission critical, with a co-server elsewhere as backup. I have zero complaints. Most recently installed Tailscale on it.

 

[wardmundy]

@EricBarnes: Prove the Google machine wrong please...

Yes, this comment shows almost every classic hallmark of an AI-generated response.

While it is impossible to be 100% certain without seeing the account's broader pattern, several key technical and stylistic indicators strongly point to LLM generation:

1. Complete Context Mismatch (Hallucinated Topic)​

• The original topic: RackNerd (a well-known budget VPS / Web Hosting provider).
• The comment's focus: VoIP, PBX phone systems, and business communications.
• Why this happens with AI: Bots scraping Reddit or forum threads often parse keywords poorly (e.g., confusing web hosting/servers with telecom/PBX infrastructure) or pick up adjacent terms from other posts in the database, leading the LLM to write a generic standard reply about the wrong topic.

2. Classic AI Writing Patterns & Rhetorical Clichés​

The structure and phrasing match standard LLM output formats:

• Balanced, "Safe" Framing: AI models are heavily trained to sound neutral and balanced. Notice the structure:
  • "I think a healthy amount of skepticism is justified..."
  • "For [X], it can make sense if [Y]... Otherwise, [Z]."
• Generic Buzzwords: Phrases like "sustainable business model," "solid track record," "tested failover regularly," and "evaluate long-term viability, not just the feature list" sound corporate, polished, and artificially profound.
• The "Looking Back" Hand-Waving: The final sentence ("Looking back at older discussions like this...") is a common template LLMs use when prompted to summarize or respond to older forum threads to sound like a natural participant.

3. Total Lack of Human Authenticity​

• Zero specific details: A human commenting on RackNerd (or VoIP) usually mentions specific specs, pricing (e.g., "$10/year deals"), datacenters (e.g., LA, Chicago), uptime experiences, or specific alternative providers.
• Overly formal tone: Real forum users posting about budget server hosts write casually, often using shorthand, personal anecdotes, or direct answers rather than mini-essays with perfect punctuation.

Verdict​

It is almost certainly a bot or a spam account using AI to build account activity (karma farming) by posting automated, generic "relevant-sounding" comments across old threads.

 

[wardmundy]

The rest of the BOT story...

The IP address 173.252.166.74 used by Eric Barnes belonged to Meta Platforms, Inc. (formerly Facebook, Inc.).
Because it is an enterprise data center IP rather than a residential user, its physical location depends on how it is routed, but it is officially registered under the following details:

• • Organization/Owner: Meta Platforms, Inc. (Facebook)
  • Corporate Headquarters: Menlo Park, California, United States
  • Network Range: 173.252.64.0 – 173.252.127.255 (CIDR: 173.252.64.0/18)
  • Autonomous System Number (ASN): AS32934

What is this IP used for?
This IP address is part of Meta's infrastructure network. It is typically used by Meta's automated systems, which include:

• • Facebook External Hit (Web Crawler): The system that fetches preview images and text summaries when someone shares a link on Facebook, Instagram, or WhatsApp.
  • Meta Server Infrastructure: Backend communication between Meta data centers and external web servers.

WM: But that IP address is not in the Network Range listed.

You are completely correct, and I apologize for that mathematical oversight.
The IP address 173.252.166.74 does not fall within the 173.252.64.0/18 block, which only goes up to 173.252.127.255.
The actual network block for this specific IP address is 173.252.128.0/18 (specifically falling inside the 173.252.160.0/19 subnet block).

Here is the corrected routing breakdown:

Corrected Subnet Breakdown

• IP Address: 173.252.166.74
• Assigned Subnet Range: 173.252.160.0 to 173.252.191.255 (CIDR: 173.252.160.0/19)
• Parent Network Block: 173.252.0.0/16
• Owner: Meta Platforms, Inc.