When are we going to have something like Affa?
I just had a server crash, and Boy the restauration is long…
Thanks,
2 Likes
At first look I thought it’s spam but I didn’t know Affa. It is an old backup solution from SME server and seems similar to the NS7 hotsync feature.
TBH, I’m not sure if it still makes sense for NS8 clusters as it was a simple master/slave solution.
I change the topic category to Feature
me too 
2 Likes
it makes me think we could have a replication of a module each X minutes, when you lost the module on the node you can use the replication on the other node
Not sure the redis database will love it but
2 Likes
The best there was, saved a lot of lives back then. Sadly the author (Michael Weinberger) went silent never to be heard from again. Luckily some guy named @stephdl stepped up to the plate back then, a young guy… 
4 Likes
Now you are speaking a language i am loving to converse
1 Like
Hi @mrmarkuz
Affa allowed us to have multiple incremental backups throughout the day. It also did daily, weekly, monthly and yearly backups.
And if a file got deleted, one could rapidly restore it.
It’s a rsync script with a command line interface.
Nothing today compares to it.
Thanks,
1 Like
Rise at any given moment !
Yes it was a fantastic piece of kit. I’ve made a simple Rsync script that backs up to offsite daily the /home folder, and hourly incremental changes. It’s okay and saved me when a Debian 13 upgrade was done on a cluster, but still very basic compared to AFFA.
The one advantage is as each file is changed, I use Rsync to keep the old versions, which means I can recover different versions of files as required. ie when a staff member overwrites a large 60 tab spreadsheet, I can roll back almost instantly to last incremental, so only lose a maximum of an hour of work.
1 Like