WhiteSites Blog
WhiteSites Blog
I just got done helping my client move 3 servers we had colocated at a Datacenter in North Houston. From what we heard our colo provider didn't pay their bill to the datacenter and as a result the datacenter pulled the plug on them. Our servers went dark last Friday on February 6th, 2026. We were able to finally pull them Tuesday, and get them colocated with Dartnode which has rackspace at TRG Datacenter which is just 30 minutes north of our old datacenter. If you have hardware with in Houston at the same datacenter as us, you will need to start looking for somewhere else to colocated it. So far Dartnode has exceeded all my expectations with helping us get back up and running quickly.
We originally colocated our servers at a Houston Datacenter back in 2017. At the time this was with Root Level Tech. A few years later my understanding is another provider took them over. The transition went well, and everything was solid for years. We of course would have the occasional need for a hard reboot, which after reaching out to our contact was usually completed within an hour or two. At one point in time we had 3 x 1U servers with them in colocation. Our bill was about $130 per U. Our network was only a /28 ( 16 IPs ) so nothing really expensive. Speed always felt fast, we were on a 1Gbit port. Never had to pay any overages for transfer, but we really didn't use that much data. Figure maybe 1 TB per month in total.
Friday, February 6, 2026 mid morning we lost connection to our servers. Naturally we thought we needed a hard reboot, but that usually would mean only 1 server was down, not both. Tried Pinging them and no response. Reached out to our contact, and they said their houston network was down, due to a routing issue. They then dragged us along Friday with the typical give us an hour or two. Night time came, and they said they were trying to find a solution to get them through, and inferred Monday they would be up. The way they talked about it sounded like a hardware issue, like a Router or Switch had gone out. This means at least 3 days where we can't get into our sytems, can't get or send emails, which is devastating to our business. Monday came and we were still down, with communication between us and our contact starting to get quiet. They stopped resonding to Texts, and wouldn't answer our calls. Granted if their network is down I assume that all their customers are likely blowing up their phones, so I didn't take it personally.
We told them we needed to pull our servers, and they delayed once again. This is a sticky situation as we are not customers of the dataceter, we are customers of the colo provider, who is a customer of the datacenter. The datacenter can't just release our servers without permission from the colo provider. We finally connected with a person at the datacenter that was going to help us get our equipment out. By Tuesday evening back and forth emails between Us, the datacenter, and our colo provider, got the datacenter's legal team a paper trail so they could release our servers. My client sat parked for 90 minutes in front of the datacenter waiting for them to finally release the servers.
I did some searching online to find another datacenter in Houston to colocate our servers with. I was hoping to keep them at the same datacenter, as that would be the easiest, but I finally came accross Dartnode, who did colocation with individual Servers. They had rack space at the TRG datacenter just 30 minutes north on I-45 in spring. The price was $65 / month per 1U and that included 1Gbit port, unmetered transfer, and a /29 of IPs ( 8 IPs, 5 Usable ), and 2 Amps of power. No setup fees, no BS. My client Drove the servers 30 minutes up north and Dartnode's tech greeted him and got his equipment into the racks. We had the servers live with all their IPs provisioned 2 hours after leaving our old datacenter. Communication was direct, no middlemen. I was able to talk with the tech on my phone while he set everything up, anything that needed my help he quickly hooked up a KVM so I could further configure the sytems. By 11PM Tuesday we were done.
Our old datacenter had an impressive facility in Houston. I have to give their support and legal staff credit for helping us quickly resolve the situation, so we could get our servers out. I have no hard feelings towards them, and would highly recommend their facility. Unfortuantely we are not big enough to get our own Half Rack, and The Datacenter doesn't deal with individual 1U server colocation, so we had to move.
They served us well for years. Remote hands was always relatively prompt. But at the end communcation fell apart. If they were having a major issue, we would have rather they be honest and inform us, so we could get our equipment moved. Our business can't go days being offline. After talking with some people more familar with the situation. What we were told is our Colo provider didn't pay their bill to the datacenter in December. And then failed again in January and February. This wasn't a hardware or network failure like we were told. This was the datacenter pulling the plug on our colo provider for not paying their bill. Granted that is only 1 side of the story. I am sure leasing a cage at the datacenter is not cheap. Granted we were paying about $130/month per 1U for colocation which turns out is 2x what we are now paying with Dartnode. You need a certain number of customers colocating just to break even, and if some of those clients don't pay their bill, it just makes it harder for the company managing the cage. My client also said he has noticed the servers getting slower over the past few months. I just assumed it was his desktop, or even maybe the SSD drives in the servers were starting to show their age. But after we moved to Dartnode things got a lot faster. Tells us it was the network, and not us. Don't know if our colo provider to save money started to try to reduce their speed, but I do know that the Thursday before we went down, when I logged into our servers, the port showed 100Mbit and not 1Gbit. So someone was throttling us. As once we moved to dartnode both servers were showing a full 1Gbit port. and speedtests showed we were definitely getting that 1Gbit. I should have run a speed test that final thursday before we went down but I didn't. I am not sure what happend with our old colo provider. Don't know if their customers weren't paying, or some kind of mismanagement lead to this.
Thus far Colocation with Dartnode has been excellent. They were quick to assist us in our time of need. Their network is fast, and their billing is very simply and direct. I have my client's 2 servers over there right now. If I ever get into a pinch with some of the other servers I manage would move them to Dartnode in a second. Only thing holding me back is the "If its not broke don't fix it" mentality, as the facility they are currently at has been solid for over 15 years. Plus I have multiple /26 blocks and redoing the DNS for accross dozens of Godaddy, Namecheap and dynadot accounts is not how I want to spend my weekend. The biggest thing I like about Dartnode is they answer their phones. If I have an issue I can call them, and talk to the guy who is physically at the datacenter to get a reboot. Not go online and submit a ticket, so that some tech I don't know gets around to it. The Personal Service is worth it, and the price can't be beat.