I am a new CF VPS Pro customer and I am having all sorts of downtime problems with my VPS. I am told that there have been several hardware failures. As a result, I haven't had one downtime-free day since last Thursday. On Friday, my site was down for hours, and over the last several days, I have had at least 30 minute to an hour outages.

My customers have been angry over the downtime, of course, and my inability to give any sort of advanced heads up. I've now told them several times that the problems are resolved, only to have them return.

Is this typical? What are the experiences that other CF VPS Pro accounts have had? I just switched over from CystalTech and I'm wondering if I made a mistake.

Can you email me (jamie@hostmysite.com) with your Hosting Account number so I can investigate this directly? This certainly isn't typical of our VPS Pro plans.

Offhand I think it may be due to hard drive issues - I know that we recently got a bad batch of Seagate drives and last night we replaced a LOT of them due to failures across the board.

Hello,

Since Jamie cannot comment about CrystalTech for reasons I forget, I should note that I too switched from CrystalTech to HMS. I've had some problems on both ends - it's an inevitable part of hosting. Yet I can say that overall, the experience at HMS is much less bumpy than at CrystalTech. It was not a mistake to move. When I was at CrystalTech, support was not as helpful in resolving issues such as this - long-term issues often remained. Once HMS gets the issue with these hard drives straightened out, which I know they will, everything should be just fine.

Yes, I am one of those affected by the drives. I was looking for additional verification, like from comprug that this is atypical. I have only been a customer for a couple of weeks, and this week has had about 12 hours of downtime or so. For two days it was pretty bad, but at least an hour for the last few days has been unavailable.

HMS has done a fantastic job on the customer support side, and I really appreciate it. However, the technical side has reassured me every day that the outages were over, only to have additional ones to follow.

Long story short, I was just hoping for some stories one way or the other from existing or former customers to help me gauge if this is something that is ongoing or a freak one-time incident.

Hi mgreider...

I'm also one of the HMS VPS Pro customers affected by this drive issue. I feel better knowing the technical details of the root cause - now makes more sense to me and understand how this has been a bit hard to track down / finally resolve on the technical side. The downtime has been tough while they work through this, but I'm confident now that once this is over we'll be on a stable / reliable box.

I've been an HMS customer for a while now and their tech support / customer service is outstanding in comparison to other hosting providers I've had. There's nothing more frustrating than having a simple issue crop up that's out of your control - and it taking days to resolve on the hosting providers end. I've been there, done that... but with HMS any issue is generally resolved right away while on the phone with them. By the way, I'd definitely suggest calling their VPS support phone number vs. email or chat. I've had mixed luck with chat and sometimes they end up having to open a ticket and pass it along to the VPS support folks. I've always had very good experience with the VPS phone support.

Here's an article from a 3rd party about the problems:

http://www.theregister.co.uk/2009/01/16/barracuda_failure_plague/

We are also on a VPS pro box. And Our site has been going down about once a day. Its only down for maybe a minute or two, but still my client is forking out big bucks to ensure that the site stays online and does not go down. I understand that sites go down, especially on shared environments, but what I want to know is why? Clients will only believe the " Act of God reason " for so long.

While I can't provide your specific answer, in general the problem typically comes back to the fact that even the high end VPS plans are not meant to be 100% uptime environments. If you want/need 100% uptime, you should ALWAYS be on a dedicated server. If you're REALLY serious about it in fact you should have multiple webservers and multiple database servers (load balanced and clustered, respectively) supporting your site(s). Unfortunately, there is a cost associated with that, so it all depends on how much the availability and redundancy is worth to you and your clients. On the low end of things if you're running an ecommerce store that makes about $100/mth then of course you should only pay a tenth of that towards hosting fees. If however you make $10,000/mth then you should expect to be paying a tenth (or more) of that towards hosting fees to make sure your business stays up and running - after all assuming an even distribution of traffic in the latter case that means you lose about $15/hour whenever your site is offline.

One other thing to mention is that it is not necessarily HMS's fault if a VPS goes down for a few minutes every day. I had a VPS a few months ago to run a site that recently closed. I didn't configure the site in the best way possible, so it went down a lot. Since a lot of things on a VPS are client-installed, it doesn't always work as well as something installed and tested by HMS.

So Jamie... based on your posting, the dedicated servers wouldn't have been plagued by the Seagate drive issues?

No, my last post was more in response to the 'down a few minutes each day' train of thought. Obviously no matter what environment you're end if you've got bad hardware then you've got bad hardware. I am assuming here that the downtime whitesites is referring to is NOT related to the Seagate issue, just general instability at times, which is quite probably due to the inherent shared nature of the environment.

Remind me to sell my stock in Seagate, before this thing goes national.

LOL, I wouldn't jump the gun just yet. Over the last decade I can remember at least three or four cases when I found the 'perfect' hard drive manufacturer, and in each case they all had a bad run or two when issues arose with a certain batch of drives.

I remember about 5 years ago we had the same thing happen with a batch of RAM. I'm not in the Infra department so I don't know where it came from, but we had one order of RAM in which every stick started failing at about the 1 year mark. It was really weird (not to say inconvenient) at the time...