Recently I conducted a 6-month trial of AtomicVPS' hosting.

Actually, it was never intended to be a trial.

It began by being a customer of AtomicVPS' parent company, HostNexus, for close to 10 years. And when I switched from HostNexus to the AtomicVPS offering, I had no intention of ever going anywhere else.

LONG-TIME CUSTOMER

I started hosting some web sites back in the early 00's. At that time, I set out to look for a solid, reliable hosting company with good customer support. After reading people's reviews and experiences -- including reviews here, I'm pretty sure -- I chose HostNexus.

I became a reseller of their shared hosting, and close to 9 years went by with few problems. The ones that did develop, were fixed. I was a happy customer.

But then a particular customer whose email was important to his business began to report problems with emails going through.

Every time I reported an issue, it was fixed. But they kept recurring. And my customer started to get frustrated.

Now in retrospect, perhaps I should've known that my reseller hosting setup wasn't appropriate for his needs, and that removing the server IP from email blacklists was never going to last, because it was all on a shared server.

But as I recall, several of my sites, including his, were on what Plesk called a "dedicated" IP. So I had the impression that their email was already going through a dedicated IP address. Therefore, it had seemed to me that delisting it from the blacklists ought to take care of the problem. I didn't realize at first that even though the web site was on a dedicated IP address, the email wasn't.

Of course, I eventually came to understand that shared hosting is fine, just as long as you don't need reliable domain email. Which, in my book, makes it pretty useless, unless you pay to host the email somewhere else.

A VPS was what I needed.

And HostNexus had a VPS wing, called AtomicVPS, that could provide that.

Laurence, the head honcho of HostNexus, assured me that moving over would be easy and painless, with no downtime.

It was a fully managed setup, with 24-hour-a-day support. Well, that's good. On the whole, I had been happy with the support at HostNexus, and it was the same team at AtomicVPS.

So I told 'em to move my account. It was a no-brainer.

ATOMICVPS RELIABILITY, MID-TO-LATE 2013:

I started on AtomicVPS in mid-2013, about a year after they launched the business. And the problems began immediately.

Over the next 6 months, I experienced the following problems that were mission-critical -- for hosting, for email on at least one domain, or for my entire account:

1. Multiple web sites not working after the transfer: "Forbidden: PHP engine is disable."

2. Ecommerce site down: It could not connect to its database.

3. Ecommerce site scaring buyers away because the SSL certificate hadn't been transferred over.

4. Email down, with message: "Can't connect to server."

5. All web sites down.

6. I couldn't log in to my control panel: "The reseller account was suspended."

7. All web sites down.

8. Main account "suspended" again, once more for no good reason.

9. Account not suspended any more, but I still couldn't log in.

10. All web sites down.

11. Web store not working.

12. Web store administration not working.

13: Multiple sites broken again, giving the error message "Forbidden: PHP engine is disable."

14: Email bouncing with message "blocked for abuse." (What? I moved to the VPS to get away from that.)

15: More bouncing email.

16: Multiple web sites down after a hardware upgrade.

17: Email down again after supposedly being fixed.

18: Multiple web sites screwed up after supposedly being fixed.

19: All web sites down. Again.

20: SSL certificate not working for web store.

21: All web sites down. At least one receiving no email, either. Probably all of them.

22: Inbound email worked, but outbound email didn't. This affected all customers using domain email.

23: Outbound email for all sites suddenly down again, 3 weeks after the last incident.

24: I discovered that the databases for one of my domains had never been transferred from the shared hosting.

25: Couldn't log into webmail on any domain that I tried.

As far as I can tell after reviewing things in detail, none of the issues listed above were caused by me; all of them were caused by some action or inaction of the hosting company. Of course they did transfer my sites over to the VPS, so that I didn't have to. But they left a couple of things out along the way.

All but the last two problems were things that I asked them to fix. The failure to transfer the databases was on an inactive domain, and I had a copy of the one database I cared about. By the time of the last issue, I was on my way out, so I just worked around it by using POP & IMAP.

In total, over 6 months, I opened 16 mission-critical tickets, as well as a few less important ones. Obviously, some of the tickets involved multiple issues. At least a few of the issues listed were caused in the middle of a ticket, as a result of attempted fixes on earlier issues that I had reported.

So in terms of reliability: For the entire 6 months that I was with them, I experienced an average of very close to 1 mission-critical failure per week. True, a few more were at the beginning. But not many. Over 6 months, the incidence really did not decrease that much.

ATOMICVPS SUPPORT, MID-TO-LATE 2013:

So how was the support, when things did fail? Well, to try and answer that question, I compiled some stats on response to the issues above.

Average replies required by me before a ticket was completed (often involving troubleshooting/ testing on my part): 4.

Average time until a mission-critical ticket was complete (usually with active participation on my part): 13.6 hours.

Or, if you divide it up into average time per mission-critical failure (24 such in 16 tickets), that would be 9 hours.

So it took 9 hours, on average, to fix a failure that was in some way mission-critical for me, or for my customers.

And this was with very active participation on my part to help the techs do their job. I usually watched the tickets about as closely as I could, and replied to questions as quickly and accurately and helpfully as I could.

To put it another way: Over 6 months, I waited the time equivalent of about 27 8-hour work days for AtomicVPS technicians to fix mission-critical problems. That's about like spending 2 months a year waiting for your serious web hosting issues to be fixed.

Except, again, I wasn't entirely just waiting. Most of the time, I was doing whatever bit I could do to help them fix the issues.

At least 7 times in 6 months, I stayed up very late at night; sometimes, literally all night, interacting with tech support trying to get a fix for a critical issue.

SOME POOR RESPONSES

There were 2 instances in which the tech seemed to have given the ticket only a cursory glance before replying --

Or, in other words, he didn't seem to have really even read the ticket.

On one instance, the tech apparently didn't bother even to check the site, when the problem was obvious.

3 times, a tech told me a problem was fixed when it wasn't.

In one memorable instance, the tech told me I needed to personally fix a problem that they had created. (And it was clear that they had created it.)

On another occasion the tech said, "Hello, I have called the script now. Please check if it is working."

But I had made clear from the information in the ticket that the problem had never been a failure of the script to execute.

On another, the tech asked me for information that I had already given earlier in the ticket.

On another, the tech told me, "Please let me know the exact steps to recreate the error from my end," when I had already given him all the information that I as the customer could POSSIBLY give him.

Finally, there was a memorable (and somewhat amusing) incident in which the tech was setting a cron job because the control panel was broken and I couldn't configure it myself. At one point I asked if he could set it to execute an hour later for me, and he informed me he couldn't... because the server was on a different time zone than the one I was in. He had earlier set it to execute at 9PM US Central time/ 10 PM Eastern time. But he informed me that he couldn't set it to execute at 10 PM US CENTRAL time, because the server was in New York.

My reply to the above was a simple: "It doesn't matter what time zone the server is in. You can always set the cron job to take place an hour later. Right? Or - to spell it out for you - 10 pm US Central time is 11 pm EDT time."

At that point, I was beyond getting frustrated or angry any more.

By the way, if anyone doubts anything, I saved the full documentation, including all of the many long conversations between myself and tech support.

The record time to fix a problem was 5 days, 1.5 hours, after 11 replies on my part, for a low priority ticket. (I didn't include that particular ticket in the list of major issues above, or in the time statistics.)

So how was the support? Well, any adjective is subjective and a personal opinion. The support was there 24 hours a day, 7 days a week. And, they did respond.

However, in my personal opinion as the person who experienced it: The support sucked.

In fact, when I began getting email reminders from them to pay the bill, I was briefly tempted to send back a series of email messages in response to their invoices that would characterize the kind of support I felt I'd received over the past 6 months. Like this:

* (Without making the payment)"This is all paid now. Please let me know if you have any more issues."

* "Hi. I got the receipt from you, thanks."

* "I am not able to pay this at the moment. You can go and pay this yourself by logging in to paypal and entering a payment to (address)@atomicvps.com."

* "I called your bank and made sure they're still in business. So everything should be fine."

* "Please let me know the amount and the payment information, so that I can get you paid."

* "I don't see the problem from here. Please let me know the exact steps I can take to see that you haven't been paid."

* "I'd like to pay you, but you're based in the UK. Unfortunately, I'm only able to pay in US dollars."

I didn't, of course. But for a moment, I was tempted.

THE BOTTOM LINE

* I appreciate the 8 or 9 good years. And they WERE good years. But the past year significantly eroded the good will that was built up.

* I think Laurence is well-intentioned.

* If you only need shared hosting, HostNexus may still work for you.

* AtomicVPS, a year after its launch, does not appear to me to be ready for prime time.

The gap is in its customer support, and in what they're doing to run the business.

Things break with no provocation. Things that (from the customer point of view) should never have broken in the first place. In my experience, they broke ALL THE TIME. And it takes way too long, and way too much customer interaction, to get them working again.

From my point of view, the problems seem massive. Do Laurence and his team have the commitment it will take to fix them? I don't know. From the outside, I seriously doubt it. But that's only my opinion, and I could certainly be wrong. Laurence frequents this forum, so I'm sure he will see this post. I will leave it all to him and his team.

I don't have any ill will for them, although I would say at this point that the bad memories outweigh the good.

But if he, and they, are capable of learning from customer experience, then maybe the whole thing will turn out to be a positive for them in the end.

As for me, I'm gone. I have moved on to other hosting. I won't name my new company; I will only say that once again I did my research and tried to aim for a company that had the best reputation when it came to customer service. I haven't been disappointed. So far, the service at the new company has been awesome.

Hopefully it'll stay that way, and I won't need to move again.