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
EPYC
Old but gold.
I did ask them for an API key once, they told me to gtfo, so seems like it hasn't changed much.
George and Mayra were the best a few years ago. When they outsourced the rest of the helpdesk to India, things took a turn. It seems those employees didn’t have any experience in the hosting business.
Can anyone here that has servers there right now say if the network still has issues? I've been looking around at some providers and a lot of them seem to be using psychz network.
no good provider is gonna be using psychz, their network is still weird after all these years. If you've ever been inside any of their facilities you'll know to stay away from anything associated with them.
can you give some concrete examples
main one is that connections keep getting dropped, it's almost impossible to have a ssh session lasting more than a few hours. makes it nearly unusable for some tasks. their ddos mitigation is also pretty touchy and will blackhole legitimate traffic, and their outsourced support is terrible at dealing with it.
Their facilities are also a mess, trash all over the floors, racks being cooled by box fans, and their remote hands are some of the laziest I have ever witnessed, having seen a few of the deployments that they've done.
How recently was this? This sounds pretty terrible. I'm surprised there's not more mention of this on these forums.
I’ve had their Ashburn dedicated server for a bit until I cancelled because their prices for additional IPs are too high. I found their ashburn network to be better than LA but still just ok. I prefer ReliableSite for low cost 1gbit unmetered dedis in US.
In terms of network issues, I haven’t experienced any issues at both LA and Ashburn
To be fair, regardless of the experiences, at least they've been in business since the early 2000's. I used them in Dallas ages ago and my experience was perfectly fine.
Well you haven't used them in a very very long time I believe so that information probably isn't of much use anymore.
You also said you didn't have any issues with tailormadeservers when they had the same network issues on 5 different prefixes I checked.
As someone who has quite a few servers with Psychz I'll weigh in a little.
They are mostly in Equinix and similar decently reputable locations. I would be very surprised if any datacenter allows you to have trash in or around your cage today. Its a fire risk and in my experience even a label thats fallen off is enough to get you a bill for removal!
BTW at-least for the initial install in Sydney they did that in house (I met with William when they did that deployment). I was actually a little impressed they didnt hire remote hands for that considering how far from their base it is. Doing the work yourself usually leads to better quality work.
They are aging.
I'm not sure if there is any big projects on the go with them. I know I'm still asking them for a few things and lately havent seen alot happening in recent times. I think they are trying to launch object and block storage but I'm not really following that. Maybe thats not going smoothly.
FYI they have an API for server ordering, some mitigation control, etc. You can get the key from the panel without human interaction. The API is not particularly new, but it did get expanded on a few years ago (including a couple things we needed).
As far as network issues I would say that their networks are generally stable at the moment. There has been times when I would not have described locations as generally stable. As a customer I am currently in the loop on the following issues with them:
If you are announcing via your own ASN in LA then Cogent recently changed their policies you may be experiencing interruption on these routes. This is why we have pulled our routes from LA (for the last month). You will need to supply Cogent with a LOA specific to Cogent to have them route your traffic (uRPF) this is only observed as a failure to egress over cogent routes. They explained this issue as the fault of Cogent as previously they did not uRPF.
If you are in London the router there badly needs an upgrade. It was supposed to happen 4th of June but aparently the parts did not arrive on time. Its supposed to now happen "early November". If you are seeing slower than expected performance in London during peak times you are likely seeing this. They did some work to resolve the worst of the issue (or maybe some heavier clients of theirs left) as the severity decreased. I havent yet requested an update (thats on me). From what I understand that network location has been at capacity for a while and they are space limited.
I don't have any known (replicated / traced) issues in other locations. We did see some issues in Chicago the other day but those seemed to be GTT at fault.
There is definately a couple people there who hold it all together and if you have done the leg work to the point you are running laps around the outsourced support then you need to get it moved up. The outsourced support should do that if for some reason they don't, request escalation.
Yes I got initially an API key, but for some reason at some point, the API keys where bound to a single ip address.
You couldn't change that anymore.
I had to setup a bloody proxy to keep using it, because they keep refusing to update the ip or give me a new api key.
The support is really cutting edge.
@Neoon I had the exact same issue. In our case because we regularly cycle our Kubernetes cluster in line with Kubernetes best practice.
1.0.0.0-255.255.255.255is a supported "IP address" you can use, its one of the things we requested. I'm not sure if its documented.Can't blame you on feeling that way. Sometimes dealing with their support can be incredibly frustrating.
We have had our growing pains in the past but I am here to help address any issues with stability. Our network has grown 10 folds where all our peers/transit are a min of 100gig and likely maxing out alot of our MX where we are looking to make those upgrades.
I was not aware of the London issue I will check in with our team on whats taking them so long to resolve.
As for the API I am not understanding this if you can get me a ticket ID via DM I can get our developers to work on that.
Our focus on development has shifted a little bit where we have been designing more cloud services / block / object storage at the moment. However that does not mean we'll let our other product line age.
There shouldn't be any drop issues if there are let me know about them and we can get them addressed. We run 16+ locations so some may have weird behavior that I am not aware of.
It appears london has had its code update (sorry not hardware, had to re-read the communication).
So likely the improvement was that. Its not perfect in the time since but much reduced impact. As I said on slack, knowing that the code upgrade is done I'll do some tracing. I had just allocated them to the known issue and assumed it was delayed.
@WilliamProfuse I think think I solved @Neoon's problem. And I suggested a documentation change to Ron to expose the configuration we use.
Talking about their LA location, they have a lot of control over that specific facility so it's a total dump. Of course Equinix etc will force them to keep their space clean.
@bobert
I've had several of their LA servers. I've had no issues with replacing a bad hard drive so I consider support very good. Their network is hit or miss. You can end up with a server with intermittent packet loss, slow speed, or otherwise excellent performance. I suppose it depends on which dc and rack your server is located. The last LA server I got from them was top notch. I currently have servers with them in Ashburn and Dallas and they are great. Excellent speeds to APAC and better than some of the highly rated providers here.
My experience at the end of 2025:
I’ve been building things on the internet for over 20 years and have worked with a wide range of hosting providers in both professional and personal capacities. In that time, I’ve never encountered a host of lower quality than Psychz.
I purchased a dedicated server from Psychz specifically for BGP functionality. Their order page advertises a maximum 72-hour provisioning window, yet my server took more than three weeks to be delivered, with no proactive communication during the delay. Their explanation was that they were out of hard drives, which raises serious concerns about how long hardware replacements would take for existing customers. When the machine finally arrived, the storage performance was among the slowest I’ve encountered in over a decade.
Once online, the network performance was unacceptable. Latency to machines in the same country routinely exceeded 200 ms, numbers more typical of transcontinental routing. Worse, the BGP session was unstable and would drop multiple times per day, often for an hour or longer. Psychz did nothing to address the instability despite multiple tickets. They repeatedly insisted the problem was on my end, even after I provided them with packet captures clearly showing the issue was occurring somewhere within their network.
The exact same BGP configuration works flawlessly with other providers, so the problem was not on my side. Between the provisioning delays, poor routing, unstable BGP sessions, and ineffective support, my experience with Psychz was extremely disappointing. I can only recommend that others avoid Psychz entirely.
haha to be honest I had forgot all about them, I do remember a thread many moons ago called something like "Is there any reason not to just blacklist Psychz entire ASN at this stage?" fun times.
Thanks for the nostalgia hit.
I had a 1230v6 in 2024 in Ashburn, had no problems, it was funny when I checked server location online it seemed like it was sitting in an open rack by its description (un moiunted by the trash
), but it worked fine and a good price. They advertise on Webhostingtalk
Did you cancel?
Who did you move to?
I have been using Psychz Dallas colo for years. Actually, it's pretty good facility (Prime Data Center), clean, security good, remote hands good. Gonna be visiting it again in a few months
From time to time, I also get their budget dedis - I can't complain. I have only used their TX dedis, usually for 3-6 month term.
ps. hivelocity has a large presence there as well (you can see their large office while walking to large colo/server rooms).
pss. I have met also a few other dudes that use Psychz for colo while working on the gear there - no complaints at all.
their non-LA colo is ok, I feel like most of the complaints come from customers in LA. I wouldn't personally use them for much but they upstream a few hosting companies that I use and I haven't had any major network issues as a result of that.
Just wanted to bump this thread. I tried out a server from tier.net which uses psychz in dallas. The server is getting random bouts of 100% packet loss for a few minutes every month. 3 instances yesterday. They are blaming it on psychz. No idea who is really to blame.
lol I remember that video, thanks for the blast from the past
Sounds like it is tiernet fault for using Psychz.
I recorded DNS being down though and 8.8.8.8 is going through de-cix and not psychz though.
We experienced nothing of the sort on the Psychz network in Dallas TX yesterday, so it is not an overall network issue at that location.
~ SMARTHOST