I've noticed that you don't have offsite backups. This can be an issue really fast. While you have separate owners for the backups which is good, you still need a "cold storage" copy, or something that doesn't instantly sync in the case of something getting on the server still.
Here's my setup:
I have my laptop(entirely encrypted), all relevant important folders and other data I want backs up to my freeNAS server using Goodsync. The freeNAS server has 4 4TB drives and can lose one without any issues. This backs up to two 8TB drives, and there's a VM on the freeNAS box that backs up to amazon cloud using goodsync again. I use boxcryptor to encrypt all the files except stuff I'm just quick transferring over the network. When I have another $200 I plan on buying an 8TB drive and putting a cold storage copy of my data 200 miles up north at my Grandmas.
I don't back up OS as I have all my configs saved and the software installation is less important to me than the 100's of gigs of data it would take to keep restorable backups that would work as I expect them to.
Currently I just use Windows backup to backup my c drive and d drive to my freenas box (network share). I am still looking for suggestions for a backup software that will backup my PC to my network share. I have tried Veeam and some others, but they do not see my network share.
Veeam isn't really a good solution unless you're running VMs, otherwise it's limited anyway. Acronis tends to do a really good job if you want to be able to restore the whole OS on a dime. I would recommend creating a new user account for your backups though, because say you get cryptolocker it could encrypt all your backups if you don't.
For your FreeNAS box. Create another user account that only has access to the backups folder. And make sure only that account has access to the backups folder.
Mirror all internal drives And have a backup drive essentially, triple redundancy
I use Robocopy to backup from the internal drives to the external. I usually run the Robocopy operation once a week, but I can run it as much as I want.
Unfortunately, there are no iterative copies of files over time.... it's just a simple backup.
This looks interesting and I may go through it tonight to finally start doing backups of mass storage to the cloud.
Mount Amazon cloud drive in Linux, and use encfs to automatically encrypt anything that goes in:
I'm going to see if I can package it up into a Docker container to easily mount it from both my NAS where I'll sync from, and desktop where I will read the files.
I spotted this thread two weeks ago and was going to reply, but didn't finish. Seems the nice forum saved what I had already written, which is a great feature. thumbup
I don't personally have too much that might need backing up, but figured I should have something around with a pinch of of actual safety over time. So a while ago I got a basic external USB housing for two 3.5" drives. It comes with its own silly little RAID mode switch, but I'm running them as individual drives paired them together as a Storage Spaces pool with parity. That enables the ReFS file system which has at least some error correction functions built in.
Since the files I store don't take up much space I can keep many old iterations. The drives are only turned on irregularly and I'm using SyncBackFree to regularly have a copy of the pertinent files ready for transfer.
Later I got me a little fanless QNAP two drive NAS that was meant for storing media, that could also be used for on the side backup. But I have yet to need it enough to actually set it up... Also it doesn't support any fancier file system.
I don't currently have drives to put in it. I am seriously considering getting some 3TB Seagate/Hitatchi enterprise drives off craigslist. If they have low hours on them. At 50 bucks each I think I am willing to take the risk. Wendell has said a few times now that enterprise drives are reliable up to the end of their lifetime, usually five years.
These are straight up enterprise drives. Seagate model number ST33000650NS and Hitachi model number HDS723030ALA640. I was going to contact the seller and ask if I could get SMART data, specifically power on hours. He's got "more than 10" available, so they are pulls from a large setup.
These are from what I would consider the more "high tech" part of the Twin Cities, the southwest suburbs. There are a lot of data centers and such down there. If they only have a year or so of time on them they would have four years of life left. By that time much larger capacity drives will be cheap new.
backblaze runs hard drives in unrealistic scenarios that the majority of people don't have to worry about. Also their numbers have been called into question years back too. In regards to Seagate there has been nothing wrong for quite some time now with their drives. I forget what models were the ones that kept failing but the problems that plagued those drives have been resolved.
Backblaze are running these drives and the data does show the failure rate to be higher on Seagate drives. The data is significant given the amount of drives they are using. You can go into the "unrealistic scenarios" argument all you want but I'd rather take the drives that can handle the unrealistic scenarios over the drives that couldn't.
What do you consider an unrealistic scenario? How does that relate to the reliability statistics they've published?
By who? When? Show me articles. Provide evidence or your assertion is bullshit. I've been watching these reports for years and never seen anything negative regarding their findings, nor the conclusions one can reasonably make based on their reports.
Further, Backblaze has skin in the game. They buy a lot of drives, and keeping good records about reliability is in their best interests. And unless you have proof that their public reports have lied about their findings, I think they are a legitimate source, and good information.
And from my own anecdotal uses of Seagate, I've had multiple failures of their drives over the years. Maybe their statistics are improving, but they are still a long way off the reliability I've seen in Backblaze's reports, and my own experience.
So kindly provide some source to back up your claims, or quit spreading unsubstantiated bullshit.
I am not going to hunt down articles just to please you. If I share anything it will be for others in the thread. Also the testing they do clearly does not reflect normal household usage. If you had a clue you would know this. Also if you don't remember when Seagate had trouble with their drives and fixed the issue it is not my problem. Finally maybe watch your tongue and you'll get a better response from me. Oh and I almost didn't catch it but now I see because you have had failures with their drives over the years you are prejudiced now against the company. Hell you probably had the faulty drives I was talking about. LOL!
It was 8 years ago when they were being bashed and rightfully so but since then they have made products that many use every single day and don't complain about. Btw if Seagate was so horrible lately why have I not seen people complaining left and right about the drives on the many sites I go on? In fact I have seen Western Digital Caviar Black hard drive complaints the last many times I have seen hard drive complaints. For the record let me state also that I have bought both Western Digital and Seagate over the years and other brands too but the latter long ago.
But you have no problem calling me out for referencing a well known study. Brilliant
What does that even mean? If you won't say, how can anyone know? Presumably, it means 24/7 operation, which a lot of people here do.
Totally clued in
Horray. Seagate fixed drives that were exceptionally shitty. Good for them. Sucks for all those people who lost data.
As far as I'm concerned you're not knowledgeable enough about anything to take a word you say seriously.
People tend to remember when they've been inconvenienced by a drive failure. That being said, I prefaced it by saying anecdotal.
Maybe. It's easy to laugh about something you found out after many users lost data due to many bad batches of seagate drives. The fact that you knew about it after the fact doesn't do anything for the miserable experience those users had.
And yet their drives are still showing up on the latest Backblaze reliability report as many times more failure prone than HGST drives.
I don't know why you want people to buy crappy seagate drives so badly. But I hope they don't listen to you.
Where did I say that HGST don't make good drives? Hell look at this info. here from you beloved Backblaze ....
Thing is though the site where I got this from goes on to talk about how Seagate are making excellent 8 TB hard drives and as for HGST uh only testing 45 hard drives and getting a 0% failure rate hardly means anything when you compare it to 5,120 hard drives and getting a 1.46% failure rate.
Also there are other interesting things in this article that speak to how Seagate has improved and how what Backblaze does isn't close to perfect in discussing how good hard drives are. Hell they don't even test enterprise hard drives. LOL!
The Backblaze failure rates are not representative of the failure rates you should expect from normal use. Backblaze use cheap consumer disks and commodity hardware instead of expensive enterprise hardware because it is cheaper for them to replace failed hardware than to use more expensive equipment. Their server boxes have dozens or hundreds of cheap disks crammed in, these disks don't have any vibration compensation or any of the other features enterprise disks have to deal with having a high number of disks in a rack. They're also binned differently to enterprise disks. So while it is interesting to see that the HGST disks hold up much better under these conditions than other brands seem to (although if these are the NAS disks this isn't that surprising) these numbers don't represent the failure rates of disks when used under normal conditions.
