The dreaded call early in the morning - 'my site is down', from a client. On going online you realize the server is down or the database connection has failed. The hosting company was doing some server maintenance at night and your website data got lost in the shuffle. Welcome to a nightmare of endless phone calls and support tickets, discussions with moronic support personnel et all.
But it does not have to be this way. A web hosting company can be a developers best friend too. After all its a symbiotic relationship.
I have been hosting sites for the past 15 years. Nowadays I get 10 times more value for the same amount of money that I got 10 years ago. Costs of servers, storage has been decreasing every year and new entrants into the hosting field are driving down costs and commoditizing the business. Margins for the hosting companies are squeezed. In fact on many products the hosting companies are losing money.
Of course the effects of this relentless squeezing of margins show up in service and support. Hosting companies are forced to hire lower quality staff and purchase cheaper equipment and servers. Ultimately the loser is the customer whose website becomes unreliable.
Hosting is a reseller driven business and many times you are buying from someone who is a third level reseller. Which means that customer service is usually slow and unorganized. While purchasing hosting try to verify if the hosting company has their own servers or are trying to resell someone else's solution.
In order to get the maximum bang for their buck, many web hosting companies cram as many websites into a single server as possible. What this means is that your website shares server bandwidth, processor time, memory and other resources with other websites. More the sites lesser your share of the resources. Try to find hosting companies that limit the number of sites on each server voluntarily to maintain good performance.
This means that the hosting company cannot keep your site down for more more than 60 minutes in any month (8.76 hours / year, to be specific). I regularly face issues where the sites are down for an entire day while the hosting team fiddles at the back trying to track down the problem.
During a recent hosting reshuffle we lost 2 months of work on a site and had to spend half a day screaming at the support team to restore it. Just like that! In the morning half the database and files missing without an explanation.
Reading the fine print (from a well respected hosting company) - 'Infrastructure downtime is measured from the time the customer opens a ticket regarding server downtime to the time the problem is resolved and the server is powered back on.' So till you report the issue the downtime is not counted no matter how long the site has been down. Once you report a problem the meter starts.
This lesson was very painfully learnt when a server crash and the hosting company had backups 2 weeks old.
This line in another service level agreement (SLA) says it all - 'You agree and acknowledge that from time-to-time the FTP Backup services may be inoperable or inaccessible for any reason'.
So basically if the server crashes and you don't have backup...you're screwed. Here the fine print can really is something to be noted.
And then - be prepared! Don't blame the web hosting company. You have been warned.
Not so cheerfully,
Ron
