The main purpose is as a faster disaster recovery solution. The secondary use case is for quick, daily backups.
About two years ago, I had a Windows 2008 Server for the fire department on which two hard drives dies almost simultaneously. HPE was nice enough to send me a couple of new drives overnight, but it didn't help the situation because the server was dead. So I spun up that entire fire department server, which had SQL running on it and a couple of databases that are necessary for dispatching fire calls and getting the information out of our dispatch system for tracking, timing, etc. I was able to bring the server up in about five minutes. I left it up and running for about two weeks.
I didn't do the bare metal restore because it was a Windows 2008 Server that was scheduled to be replaced that year anyway. I just let it run for a while and when I got new hardware I built it up as a Windows 2012 Server, at the time. I was able to do it on my timeline, rather than being in a panic situation and having to get the server back up. The entire time, while it was running in this virtual host, it was also backing up at the same time. If necessary, I could have gone to any of my backups, which sounds weird — it's a backup of a machine that's running on a machine — but I could have gone to any of those backups. That was the only instance that I've had to rely on any of its features, above and beyond just restoring files when people mess up and delete something.
The case of somebody overwriting or deleting something by accident, where I have had to recover a file or data, happens more often than I would like to admit. I find myself having to restore a one-off file about once a month. It happens more around budget time. People take the stuff from last year that they think they made copies of and make them into blank documents for this year. Inadvertently, they're working on the wrong ones and I get to restore those. That seems to happen at least once or twice every budget year.
Another scenario is that somebody comes to me and says, "This file was supposed to be in this directory. I either accidentally overwrote it or I accidentally deleted it a few weeks ago and I need it back." If they're really good about it, they'll tell me the name of a file or a directory, which gives me something to work with. Hopefully, they'll give me a rough time that they know it existed. I'll just start working through snapshots. I'll open up AD servers that have all of my file shares on them and I'll pick a date. I'll start with January first. If the file is not there I'll move to January second. I'll continue going through all of my snapshots so that I get them the absolute latest and greatest one that there is and which is still functional.
Another instance where I'm having to restore on a somewhat regular basis is when people leave, depending on the situation and why they left. They may not be doing it maliciously or they may be trying to either cover their own tracks or trying to make it difficult for the next person. They'll delete everything that they've got. I'll go into their Outlook and they've got three emails, but they've been here for four years. So I'll have to restore all of their mailbox from a few weeks prior to their putting in their request to leave, and start restoring files as well. That happens about twice a year where I have to go to that extreme.
There isn't a cut and dry process in that situation. When I get notification that somebody is leaving, I back up their email, for records retention. If I realize all their mail is gone, or it seems like there's stuff missing, I'll start restoring old stuff from the past and see what they got rid of. It makes the life of the person who will be following them in that job position about a million times easier if they've got some idea of the communications the previous person had or any notes or documents they had for that job.
I also benefit from the solution's deduplication, especially with their new software release, and how they're arranging the storage for the disaster of virtual machines. It is handled differently. It doubled the amount of space that's available for my backups. The deduplication helps, obviously, because I'm not having to back up the same caches over and over again.
