This blog and hundreds of thousands of thousands of other websites went down yesterday due to unacceptable server maintenance problems in a Provo, Utah data center. Bluehost, HostGator, and HostMonster are divisions of of Endurance International Group.
Our datacenter in Provo, Utah experienced an outage that impacted many of our Bluehost, HostGator, HostMonster and Just Host customers, causing intermittent connectivity to your sites and our internal support and phone systems throughout the day.
The problem occurred during routine server maintenance when a hardware failure quickly cascaded throughout the network.
Yet some hosting companies virtually never go down while others, like BlueHost, have continual and sporadic downtime. (This is a major reason my blogs are no longer on BlueHost. I use Laughing Squid which co-locates everything in the Rackspace Cloud. They just don’t have outages.) A solitary hardware problem should not take down three major hosting companies. That it did indicates a serious lack of redundancy with their hardware being maxed out continually.
Comments on Mashable are illuminating.
The real issue is EIG. They have taken Bluehost and modeled all the other companies after them. Which would have been great if it was the Bluehost back in 2010 before EIG purchased them. However, since EIG has purchased Bluehost, EIG’s corporate greed and lack of responsible leadership has deteriorated all these brands, with the exception of Justhost. That place sucked worse before EIG bought it.
According to Wikipedia, the current shareholders are two equity investors: Warburg-Pincus and Goldman Sachs. they paid $975m for the EIG group of hosting companies.
It also appears that redundancy is not built into the wider EIG portofilio meaning that the mirrored boxes are not available in other data-centres. Rather, they appear to have a distributed architecture with boxes (and networks) and you either land up using a box in Utah or somewhere else.