I really feel I need to congratulate the Plone community for all their effort over the years in creating what I think is a real jewel of the open source community. After a week of messing about with it, I can really say that Plone is what I have been looking for my Intranet for a while (why didn’t I come across it before?).
One of the biggest concerns I had was backup strategy and portability of the database. Well, it turns out that this is really a no brainer although I struggled to get this info out of the documentation, so I had to trial and error.
This is what I did:
a) I installed the same version of Plone on two separate machines.
b) I created exactly the same site on both.
c) I populated the site with documents and pages on one site (machine 1)
d) I ensured that the Plone instances on both servers were down
e) I copied the “filestorage” and “blobstorage” folders from machine 1 to machine 2 into the respective “var” folder
g) I fired up the Plone instance on machine 2 and voila! It is a specular copy of machine 1. It couldn’t be simpler …
This makes me wonder why would one bother using the Plone internal backup mechanism when as long as you regularly backup our var folder, you are fine. I guess the advantage is that when you have a huge database, you don’t want to have several daily copies of the var folders which would use huge amounts of space? I guess the internal backup process uses a progressive backup mechanism?
Also, I am assuming that this approach only works when the tow machines have the same version of Plone. You would not be able to use this approach between different versions as I guess the database structure may change between versions?
Anyway, I thought I would share my experience so that if other run into Plone and have the same concerns I had at the start, this may help them address their concerns.
PS. What I forgot to say was that machine 1 is 32 bit while machine 2 is 64 bit. I guess that doesn’t make a difference as I’m not moving binaries and libraries between the two machines …