Hi all,
Jim Croft wrote:
<snip> 1. Building mission-critical stuff around a single point of failure without distributed replicated redundancy is amateurish, ultimately doomed to failure, and I am amazed that everybody does it. The drive for an easy solution with smart response times wins every time. IMO, the Australia's Virtual Herbarium took a step backward when it moved from a distributed to a cached solution without building in fail-over redundancy. Yes, the new version is quicker but when it does not work you are clean out of luck. If we are going to build anything that is going to become mission-critical and expect people to use it, then I want more than one of them. </snip>
Actually I believe your criticism is only partly correct. Centralising a copy of, and indexes of, distributed resources (i.e. a hybrid distributed+centralised model) is a good thing in some crucial respects - (1) speeds up query time and also (2) eliminates the sometimes patchy real time connection issues to remote servers. In addition one can hopefully (3) centralise the copies in well supported and maintained location/s with economies of scale for content maintenance, QA, and IT support e.g. 24/7 failover service. The residual problems are then (as you correctly say) mirroring (duplicate central caches/indexes in alternate locations) plus (as you do not say) synchronisation (ensuring that changes at the remote points-of-truth are rapidly propagated), however a rollback to the completely distributed model is in my view not the answer.
Just my 2 cents worth,
- Tony