Just doing some research, is anyone here on WHT looking for Hot-Failover (redundant) solutions. A managed solution in which if one server fails the secondary backup server will take over all tasks including MySQL, Apache, dns, and file storage. Until a point at which the main server comes back online in which case the backup server replicates its data back over.

This could also be used for load balancing.

Anyone have any thoughts?

Us. Count us in. How much, specs, etc. PM me if you'd like.

who is "us" ? :)

What are you looking for exactly? Clients interested in a redundant solution? Or people to share a server? I think we need more information about your request.

Originally posted by atchoooo
What are you looking for exactly? Clients interested in a redundant solution? Or people to share a server? I think we need more information about your request.


correct please specify more information.


thank you.

I apologize for the ambiquous answer.

My company runs e-commerce solutions hosting. It's a CGI-based shopping cart---thoroughly tested---5th generation at this point.

We need a hosting solution that will not only provide website redundancy but data redundancy. If one server goes down, we want our clients' websites, emails, and services running without dropping a beat. Bandwidth or load balancing is not an issue with us just yet.

madspeed:
1.) setup (my)sql replication and put all your data into the database
2.) you'll then have a master db and x slave dbs and y web servers
3.) it's possible to automatically change the master DB in case it breaks and 'appoint' one of the slaves to become master
4.) web server redundancy by hardware load balancing or dynamic DNS (works great here with some 100M+ page views/mo)

then there are also linux clusters available... never used them though.

I need to look into a redundency solution myself....could use some pointers and hints on what to do and what to look out for.

I'm just researching to see if theirs any demand for managed redundant solutions or even smaller per-website redundancy.

interesting idea. I guess if you kept a low TTL on the dns you could fwd all requests to other ips when problems. Was thinking of doing this myself so I guess I will have to look into it more myself as well

Originally posted by f0urtyfive
I'm just researching to see if theirs any demand for managed redundant solutions or even smaller per-website redundancy.

There should always be a demand for such things :)
Where can I learn more?