You're going to have to invent a new backronym where TITSUP is usual performance.
Heart Internet goes TITSUP again
UK-based web host Heart Internet has restored service to customers whose email has been titsup since Monday. Users were unable to use the service due to a single server failure. One reader got in touch to complain: "Apparently in a modern environment, hosted email clustering means only having one server for any individual …
COMMENTS
-
Wednesday 27th July 2016 12:31 GMT Pascal Monett
Hosting is done on one server
Cloud is still in its infancy, obviously. It would seem that some cloud services are more pared to the bone than others.
The only good thing about all this is that the great unwashed will learn, by trial and painful error, that price is not the only thing to take into account when signing up for a cloud service.
Like the automotive industry, the only good thing in this evolution is that, in the end, cloud services will perform better, because those that don't will wither and die. The only question is : how long will it take ? With the amount of people who pay and find shoddy performance good enough, it might take a while.
Now, if you're basing your business on it, I'm sorry but it seems to me that you should reevaluate your ROI criteria on this one.
-
Wednesday 27th July 2016 14:56 GMT Mayhem
Re: Hosting is done on one server
Yeah, we've had enough and are now actively moving our customers off them.
They were good, but have frantically bought customers from everywhere in a race to the bottom and are frankly beyond rubbish now.
This was the harder to defend than a complete failure - why is HIS email down when MY email is fine?
-
Wednesday 27th July 2016 23:52 GMT Anonymous Coward
Cloud is still in its infancy.
> Cloud is still in its infancy, obviously. It would seem that some cloud services are more pared to the bone than others, Pascal Monett
It's the people selling 'cloud' that have created unrealistic expectations. If it ain't cheaper and more reliable than sharing a rack-mounted PC in some server farm, then what's the point of moving to the 'cloud'. And cloud had been in use commercially since at least 2006, hardly in its infancy. What's never mentioned is the amount of bandwidth you would need to tap into your remote cloud IT infrastructure and what would happen to your business if-and-when it fails.
-
Wednesday 27th July 2016 13:21 GMT Vince
"Heart describes itself as "The UK's best web hosting backed by free 24 x 7 support"."
Self-Described and baseless claim (how is that validated precisely?)
Oh...
And they didn't say "best e-mail hosting" - just web hosting... so maybe they're the UK's worst e-mail hosting provider backed by free 24 x 7 support - although I imagine others will have views on who qualifies for the "worst" accolade.
-
-
-
Wednesday 27th July 2016 19:11 GMT defiler
Re: Windows
Microsoft Exchange service provider licensing is per mailbox, so no huge penalty there for replication.
If they have any sense they'll be using Windows Datacentre which is licensed per CPU socket regardless of number of virtual machines.
If you're hosting an Exchange service you *start* with 2x CAS, 2x Edge and 2x Mailbox servers, and you run them distributed across at least two physical hosts.
It's not hard if you plan it.
-
Thursday 28th July 2016 05:48 GMT Anonymous Coward
Re: Windows
Perhaps somebody needs to tell them how to configure Database Availability Groups (unless of course they are hosting Exchange earlier than 2010).
Also, you shouldn't have 2xCAS and 2xMailbox. Best practise is to have CAS and MB roles on the same servers. Check the preferred architecture for Exchange 2013:
https://blogs.technet.microsoft.com/exchange/2014/04/21/the-preferred-architecture/
If you move to Exchange 2016, you can't have 2xCAS and 2xMB. The CAS role no longer exists.
-
-
-
-
-
Thursday 28th July 2016 15:59 GMT wobbly1
Not just Tech Support a problem for Heart
Took 5 months of constant reports to them for one of their clients to stop spamming me . He ignored emails, Heart didn't seem to think enforcing their Ts and Cs was their job. Probably their support staff were fully employed bicycling on the static bike with a hub dynamo to provide the power to run their server.