-3

I've been reading about full redundancy to achieve almost if not 100% uptime for your mission critical applications (handling financial data) running in a dedicated server, but I still can't wrap my head on how I could do this.

1) Let's say I have multiple web apps (with large and frequent files and DB updates) running in my current server. From what I read I would need to get another dedicated server (as a backup server) where I could mirror in all the files and DB data, and if my primary server went down for any reason my backup server would carry the load immediately. The transfer from primary server to backup server is done through DNS. Is this correct? Is there a more better way to achieve redundancy than this?

2) If we did the what I mentioned in #1, what would happen if the primary server went online again? How would the primary server catch up from all the changes done on the backup server?

3) Is there a definite guide out there I could read on the best way I could do this? I've searched through Google already but what I mostly encounter would not work for us because our web apps can't be down for more than 5 minutes at least and we can't lose 5 minutes worth of data.

4) Is there a service out there that could do this for us? Since we already have an existing dedicated server, the service would have to work around existing working environment. But I would prefer if I could learn how I could do this myself.

I would greatly appreciate it if you guys could point me to the right direction here.

4

This depends entirely on the architecture and availability options of the specific applications in use.

Does your database support a clustered deployment? A mirrored/replicated deployment? Both? What are the caveats of each? What's the failover/recovery process? The resource allocation requirements? What about for your web app? Can that be clustered? Load balanced? Both? What about your ISP? Do you have redundant internet connections? From different conduits coming into different sides of the building so that construction won't clobber both? What is your power Config? Do you have redundant power feeds? Redundant UPSes? A generator for the Datacenter?

These questions aren't meant to be answered here - it's meant to show you what the tip of the iceberg looks like for designing this type of solution. 100% uptime is an unattainable goal, but if you want to get as close as possible, you should hire someone that has experience doing it, as you're nowhere near ready based on the content of your question.

| improve this answer | |
1

MDMarra is correct, this is not a question that can reasonably be answered definitively here (+1).

You need to take a look at your application across all seven layers of the OSI model and then add redundancy at each level. You will most likely need input from all of the developers and admins involved.

When management starts complaining about the price tag, be prepared to have a discussion regarding cost of downtime vs redundancy. A good rule of thumb is that each decimal point (99.9% -> 99.99% -> etc) equates to another zero on the end of the cost so it will likely come down to a well researched design and execution that meets the technical needs and management's cost concerns. It makes very poor sense to build a million dollar solution to protect a $1000 application.

However, if management is writing you a blank check, I'm sure there are plenty of folks that would be willing to consult =)

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.