So I haven’t seen anything on here but thought I’d open a discussion on disaster recovery to see what everyone is using and how. Backups are very important and using the built in database tools are very good but that’s just one piece. For instance do you have a backup for windows?.. Is it onsite and off site? I use an array of technology to stay up and running. First off I use a server that is setup with a raid 10 for hard drive redundancy and speed. I also have samba perform a database export daily… Then I use a product called shadow protect for incremental backups every 15min that are vss aware which also works in tandam with a program called image manager which verifies the backups and consolidates them and also replicates them to a external hard drive. I also have a sync that happens which sends the backups off site to a PC at my home so if something ever happens I have multiple backups. I also have my samba database backups going to my 1TB one drive folder which is synced up to the cloud for another layer of redundancy. The shadow protect backups that I use can also be easily converted to vhdx files which would allow me to get my system up and going within minutes on anything that is running Microsoft’s hypervisor. What is everyone else doing for Disaster recovery?
Very nice setup! I simply just backup to an external jump drive for now. I am not too worried about the jump drive failing they have a very solid reliability rate. I keep backups on the PC itself and copies on the jump drive. The jump drive is plugged into the same server.
My next upgrade will probably happen soon but I plan to install a NAS for backups and it will power a media server as well.
If my internet was faster and more reliable I would probably use ONE Drive as well. But the internet in the small town that I live is very slow.
Nice!.. Yea my system is a little overboard but its for my parents restaurant. Since I live a few hours away I have a few ways to get them back up and running when they are in a pinch… I also forgot to mention that we I have all the backups going to my house so if I need to I can quickly get the backups up and running in a VM and just have them use rdp on all the terminals onsite to connect back and continue working with minimal down time. Its also very important to do a test restore at least once a year to make sure all your backups are good and working correctly.
LOL, maybe just a bit… but it doesn’t hurt to cover more angles than just 1 - you can’t be too careful 
My DB backup is done on Application Ending Event using a BAT file.
SQLCMD does the backup
RAR compresses it heavily.
COPY sends the RAR file to the NAS.
DELETE destroys the BAK file.
Then I periodically copy the RAR to my home system via FTP.
I have no full system backup of the OS etc. If it goes down, I’m screwed, and will need to rebuild it from scratch, except for the DB. That said, I have a completely working Dev machine that I could use in an instant after restoring the most current backup, so my actual overall downtime would be negligible.
I have samba send backups to a USB device thats screwed under my counter top so that IF the till was stolen then I would still have the backups with the gift certificate codes and amounts on.
I spose i should really set it to backup externally too maybe on my business website
Matt
Is anyone out there using a trigger to do database backup every few minutes?.. If so do you notice any slow down?.. Only reason I could see to have database backup every few min is to not loose orders or transactions.
Every few minutes seems excessive.
@kendash I agree that a full backup ever few min can be very excessive and slow if the database is large… But what about differential backups… SQL server supports this natively. I’m not sure how the samba database tool creates backups but if it could support doing a differential every 15 min that could be a realistic option. The thing is that I worry about is if the server goes down if you are using kitchen displays and not paper tickets you could be up the creek without a paddle. Everything is raided but I have seen multiple disks fail before at the worst possible time… Just trying to spark some kind of debate.
Honestly if that is the worry then the smartest approach would be to print tickets along with the display. Reason is you wouldnt have to wait to reload a backup etc. And even if you were running incremental backups… busy hours tickets happen much faster and its not realistic to wait to reload a backup and then pull those tickets back up if it even caught them all. Your cooks would simply look at the alternative printed ticket and there would be ZERO loss of time and less customer frustration. We all know when it comes to kitchen to customer… time is everything.