Man these guys drive me nuts.

They talk about 7x24 support on their website, but what that appears to mean is that you can log the call at any time in Wonderdesk, not get your problem fixed.

They apparently don't work weekends. This must be the 10th weekend since March when I've been down for more than 4 hours for lack of a support model that really covers weekends.

The http daemon on the shared server I'm on apparently died at 00:25 this morning. Yes, it's now 13:30, and all the websites on that server have been down for 13 hours. I have two domains on that server, and both are down. A typical Sunday generates most of our revenue for the week. Today will generate nothing.

The 99.99% uptime guarantee will not cover this - it falls between the cracks - so I'm just screwed until they come back in tomorrow.

I've got to change hosts. This crap is driving me nuts.

-t

What's your website URL?

An update.

I just received a reply to my support ticket. I'll paraphrase:

"The servers are up. The network upgrade resulted in little downtime. Some services are not fully operational at this time, but we're working on it."

The servers are up, but the http daemon is not running, so I can telnet to the server, but my website is down.

Hellooooo. This means that the server is down as far as my users are concerned!

-t

I've never had much of a problem with support, I always do it over phone so problems are fixed faster. They aren't there as far as phone support goes on weekends, but I always get my support tickets completed on weekends. "Jack" doesn't seem to be the brightest, but Ross and Mterry are both good at getting things done.
As for the downtime, I had 4 1/2 hours of downtime during their upgrade period a couple days ago and once it was back up everything was working fine (I'm on their Corona server).
They also seem to get things done better when you call in. For some reason if you know what is wrong and you tell them in a support ticket, they tend to ignore the fact that you told them what is wrong and it sometimes ends up not being fixed. I've never had that problem phoning in, so try that if it's a weekday.

Well, just to be fair, I'll close the loop here.

The http daemon was restarted at 13:46 according to both my access log files.

My sites were offline for 13 hours and 20 minutes because a daemon process had died - though I expect I'll never know anything more about the problem.

By the way, I've not been notified that things are working again. My call was closed with the reply at 14:18 telling me that the server was up but some things were not working...

Hosts wishing to provide stellar service without a huge amount of extra work should all have critical process monitors in place with restart code or at the very least, notification code.

I run a manufacturing system datacenter for a Fortune 500 company, and we are required to have all our systems up and available 7x24. If a critical process dies, we know about it immediately. If it cannot be restarted automatically, someone is notified and the processes are tended to immediately.

Why is this such a hard concept to grasp? Why are web hosts not doing these things? They are not expensive to do and the customer is so much happier when things are working...

-t

If you are having so many problems since March and they can't help you, then maybe it's time to look for other hosts.

Originally posted by Palm
If you are having so many problems since March and they can't help you, then maybe it's time to look for other hosts. I know. It's just that I hate to throw in the towel - but I'm getting ready to start the process of searching again.

The sad thing is it's not that I'm having any problems that I need their help on, I simply need them to keep their servers running! I'm not asking them for anything other than uptime of the server and services that they provide. I wish I had a provider that I could actually make requests of instead of simply asking them to keep their systems running.

-t

I've been loyal to Liquid Web for a while.

I've got to say that tons of the problems are from their NOC and not the guys at Liquid Web.

I thought the comment on Jack was slightly right on ( just have to get on his "geek" side ). Ross is the best because he's always thinking his job is on the line so he get's it done right away.

Once you know how to deal with liquid web, then your down time will be minimal. Remmember, you most likely looked around before you choose them and your paying one hell of a good rate for the bandwidth and server space.

I know some things about liquid web and if i was in your shoes I would still stick around ( it should be getting a lot better).

-Michael

So Michael, how would you suggest I handle the case where something breaks on a weekend, and there is no one home? The 800 number tells me to use the website. The website reports go into a black hole until someone - usually Ross on the weekend at odd hours - gets to it. Never 7x24. 5x10 with the occasional weekend hour or two is more my experience. Maybe I'm just unlucky. Maybe I'm asking too much for my $300/year (actually $600 - I have two accounts on the same server! How dumb is that!)

My history of problems with them are mostly around services failing on their servers on the weekends, and them not responding to support requests for many, many hours. In some cases, I've reported a problem like the mySQL file area being full on a Friday night, and not had it resolved until Monday morning! This is absolutely not acceptable, and is certainly not related to problems with their NOC. This is a simply server management issue. Nothing more, nothing less.

They are unable to get Akopia Interchange running after many, many requests on my part. I've given up and am looking for alternative shopping carts, and it's too bad, because I think that Interchange will work for me. I'm paying for it as part of my monthly fee, however I will never know if it will work for my needs as long as I'm with Liquidweb - as they cannot seem to keep the daemon process running long enough for me to test it thoroughly. Again, not a problem with the NOC.

This last problem was another one typical of a non-response. The http daemon died at 00:25. I did not notice until 11:20 since I was busy all morning with my family on Sunday. I reported it at 11:27. The problem was acknowledged at 14:18 and the server was finally up about 15:45.

The web services were unavailable for 15 hours, and who knows how much longer they would have been down if I did not happen to check the site on a Sunday. This should have been caught by their own monitoring software and handled long before this. A customer should not have to report something as fundamental as the web server process not running on a server. I have no idea what time the problem was first reported and by whom. There are dozens of accounts on that server. They were all offline for 15 hours on Sunday. Again, not the NOC's problem.

I hope you can understand my frustration. I left my last host after 5 years when their service levels dropped significantly. I will not wait that long again, however I am a patient man if I see things improving. I also understand mistakes - God knows I make my fair share. Repeated failings of the system however point to lack of understanding of how to support their business.

From my records I have suffered extended outages of 6+ hours for non-scheduled reasons 9 times since March. There have been a number of scheduled outages, but I'm OK with that. They have handled their scheduled downtime fairly well.

Am I asking too much to expect them to keep their servers running without me having to tell them they are broken? In one case, the outage lasted from Friday at 20:00 to Monday at 08:20. I lost the entire weekend because of a runaway process that was eating up space on /var, and ultimately corrupting the mySQL databases on the server.

If you would care to share with me privately what you believe will make things much better in the near future, perhaps that will keep me from moving my accounts to another provider within the next 30 days. I am shopping hard now, and as painful as it is to move, I will do so if things do not improve.

As for the price for the service, there are plenty of other vendors offering similar prices for the specific server pieces and bandwidth I need. It's one thing however to offer - quite another to deliver. I expect the servers to be up and available, and if there is an uptime guarantee for it to be honored. Liquidweb has managed to sneak out from under their uptime guarantee by making statements like "Even though the mySQL databases were unavailable, the server was still up so technically we weren't down," which of course does me no good for a site that is database driven.

-t

Originally posted by thewitt
statements like "Even though the mySQL databases were unavailable, the server was still up so technically we weren't down," which of course does me no good for a site that is database driven.

LOl, that's really funny. This sounds unhonest to me.