I've got a shared hosting account with Gearhost. Everythings been fine up until yesterday. Yesterday their server was down for approx. 6 hours (I guess more) and then this morning when I check it's down again. It's killing my newly set up web business. I send support tickets and each time - no apology, just directing me to look at the network status page, which tells me nothing. I've even tried calling (from Japan!) but always get the recorded message).

I don't know how the 99% uptime thing works: over a day/week/month/year? Am I expecting too much?

Hi

No, you are not expecting too much, 6 hours of downtime without any explanation for the host or any advanced warning is totally unacceptable, if downtime continues to be long and if you are not satisfied with the support and hosting, I suggest you try another host, to save you trouble in the future.

agreed - downtime of that amount w/out explanation would make me leave that host.

leojez,

The downtime was due to a webserver that went out due to a disk failure which is one of the hardest things to replace. The server was up about 2 hours after the crash but has been a bit up/down the whole day due to issues with the restore but everything is getting ironed out. I apologize that the downtime has resulted in frustration, it would get to me to. As per the emails sent we are keeping a good communication bridge between our status'. Email me personally if you need anything further.

Originally posted by xnet
agreed - downtime of that amount w/out explanation would make me leave that host.

I agree but as we've emailed and posted on our site we keep a good communication bridge with our customers. Everyone, one way or another knew when and why, thus our online network status page.

Are you saying that Downtime is acceptable if its known about???? :eek: :eek2: :eek3: :eek4:

Are you saying that Downtime is acceptable if its known about????

I'm sure he isn't inferring that. Downtime of any kind is bad, but sometimes it is unavoidable.

Originally posted by bombino


I'm sure he isn't inferring that. Downtime of any kind is bad, but sometimes it is unavoidable.

I agree with bombino on this one..

I'd also like to add that if you have a guaranteed uptime you should do your best to support that. If you're having problems-which almost everyone does-then tell your clients what's going on. They are paying you for a service and if that service isn't there they want to know! That's my $.2 though. :stickout

Well put Matt and Bombino, downtime, as unwanted as it is can hardly ever be known. In our case a disk failure will creep out every once and a while, a norm in the business. Like you said though Matt, keep the communication open and letting your customer know right away is the best way to handle these issues. In our case a broadcast email is sent to all users on a specific web server, updates are made to the site, call center is made aware of the problem to relay status to calling customers.

Thank you for replying, ryank. I understand that problems from time to time are unaviodable and can happy to any host. However, it was the lack of communication that really annoyed me - leaving me confused and not knowing what would happen.

Ok - the network status page says this: "Oberon is currently off-line while a disk rebuild takes place due to a disk failure. ETR is 45 min."

Well, to a general user (ie not a technical person) Oberon means nothing to me. As far as I knew my server has a totally different name. Also the 45 mins ETR is totally wrong. As for the emails you sent out - what emails? I received nothing except replies to my tickets saying see the network status.

If I had received an email, or your homepage had a message saying exactly what you said in your reply to this thread I would have understood and not have been left in the dark.

Anyway, thanks for the reply - NOW I know what is going on.

leojez if you're still not sure about GH's uptime guarantee it's available here. (http://www.gearhost.com/aboutus/legal/guarantees/)

BTW do you have a final calculation of the total amount of downtime you had yesterday? After having my site down over 15 hours just wondering how badly others were hit.

I am on the Oberon server at gearhost as well and I never received any emails on the subject. I only received many email from internetseer (site monitoring company) saying when and how long my site was down. Right now my sites are down with no explaination and the Network Status page on their site says nothing about this. I too do business through the back end of my site and this makes me look like a chump. My clients can't see the status of their project and can't view and pay invoices. This sucks. Answer the phone for **** sakes. I think I've been patient up until now but almost one week later and still the same'o same'o. <<Removed: support request, please email/PM directly>>

Read http://www.webhostingtalk.com/showthread.php?s=&threadid=38098

I am very surprised. I would expect that any host would be running RAID. Even more so with a 99% uptime guarantee.

Disk faults do occur and you should protect yourself from the extra work (restores, rebuilds) and damage, both financially and to reputation by a simple proven strategy.

I cannot see an argument for not running RAID. I am sure the time that was spent rebuilding/restoring would have paid for a second drive and controller.

TITAN,

We do run RAID, infact RAID 5 on all web servers and RAID 1 on all Application Servers but you must not have much experience in running RAID because even though the RAID controller can rebuild the drive it takes a while plus it takes even longer if the machine is performing actions thus we take the machine offline and let it do it's thing for 45 min. or so. After that we reset all the information and the box is back up. The server never had to be restored from backup.

Apologies, I read your comment above :

" The downtime was due to a webserver that went out due to a disk failure which is one of the hardest things to replace. The server was up about 2 hours after the crash but has been a bit up/down the whole day due to issues with the restore but everything is getting ironed out."

so please forgive me for mentioning there was a restore when clearly I misread the above paragraph.

You are correct - I know very little about RAID. My 5 years experience laying out filesystems on Sun E450's, Compaq Proliant servers and small fileservers using combinations of RAID0+1, RAID5 and RAID1 have not helped me. I must have done it all wrong though since we never took machines offline - they happily rebuilt a swapped drive in the background whilst continuing their functions (mostly i/o intensive Oracle db servers).

I must just be lucky I guess.