Cloud hosting and single hardware point of failure?
        Posted  
        
            by 
                PeterB
            
        on Server Fault
        
        See other posts from Server Fault
        
            or by PeterB
        
        
        
        Published on 2012-09-06T08:21:57Z
        Indexed on 
            2012/09/06
            9:40 UTC
        
        
        Read the original article
        Hit count: 260
        
From talking to sales I thought Rackspace Cloud was running on a SAN and compute nodes (as VMWare's offerings do), only to find out it doesn't, so when the host server goes down for maintenance all cloud servers on the server go down (in our case for 2.5 hours). I understand Amazon EC2 also has this single-server point of failure.
- Which cloud hosting solutions don't rely on a single server? I've yet to find a list by architecture
- Is there a term that distinguishes between these types of 'cloud'? Is one of these 'grid computing' and the other 'virtualisation'?
- Can a SAN backed solution provide the same reliability as 2 mirrored cloud servers on (say) Rackspace Cloud?
I am more familiar with the VMWare architecture and would like to understand the advantages and disadvantages of each approach.
I understand the standard architecture is to have multiple cloud servers and mirrored data between them; until we need multiple database servers I'm wondering if a SAN/node hosting solution would provide the lack of downtime we need without the added complexity.
© Server Fault or respective owner