Imprimer

why?

One day, due to the breakdown of the maionboard of the SME without any warning, I had to use suddenly such a decentralized network in a productiv way.

After it was clear that the hardware of the SME was broken, I tried to run the backup server as a production server by using "affa --rise". Unfornately, because of Murphy's law, it didn't work as expected this time (inspite I had positivly tested previously the conversion of the server), about 1/3 of the services couldn't be used.
I restored by hand in a dirty way the main fonctions of the SME server to get the time to build something different.
It was for me clear that "all-in-one" is too risky: "loosing one" can result in "loosing all" and this produces .......some stress!

How?

Steps:

The general concept:

The general concept is in fact very simple: (diagram coming soon!!)


Architecture:

Distributions:

I tried to use the "best" (in my point of view!) specific distribution to achieve each group of fonctions.

Catégorie : Informatique
Joomla3 Appliance - Powered by TurnKey Linux