infrastructure architecture issues
outlining the context
I'm thinking that the basic problem we're trying to solve is
this:
Create a recipe for setting up a reliable socialtools server as a
single low-cost PC on an ADSL line in someone's home.
This problem imposes certain restrictions on us:
- Because cost is an issue, we can't rely on expensive hardware.
- Because it's on an ADSL line, we can't use a lot of bandwidth, e.g. for full backups over the network.
- Because it's in someone's home, it has to be very low-maintenance. We can make sure that someone can always get to the machine in an emergency, but we can't require frequent manual maintenance.
- Because it's a single PC, we can't use a second PC for backups.
Let's consider rabelais to be an experiment in solving this
problem. Once we have the recipe, we can reproduce it with lots
of other servers all over the world.
to do
our
only task in here is to list infrastructure issues to solve.
among those issues could be:
- (1) How we'll provide failover (switching DNS to another machine
if one server becomes unreachable)
- (2) Automated backups
- (3) How we'll add new servers, located elsewhere
once we think all isssues are in here, or at least all that we can think of right now, we move them to solutions page:
InfArchitect (next task in infrastructure section on
ProjectPriorities), where we work on findind solutions to individual problems
done
context has been outlined.
ongoing
put here issues that are temporarily resolved from list of priorities, but that are ongoing tasks i.e. they will come back to LOP (list of prirorities) at some later stage. for example, an ongoing subject could be what was previously written as "Think about
http://www.infrastructures.org. Any ideas we can use?" - we could call this
insfrastructure ideas?.
dependencies
mirroring servers?
to top