Any Computer Enthusiast in the house?

[Chandu]

No, ShadowProtect has a verify option that can scan a backup to ensure that it not corrupted. I don’t know enough to say how they do it, but I verify them all, and the times I have had to restore from a backup have been flawless. The same can’t be said of the DAT backups we used to do.

Cool. Just saying test restores periodically and document it. That's a nice CYA so if a necessary restore doesn't work, you can point at the fact that you've done the very same methodology and it worked. The whole reason for the back up is that it may be needed to recover from so be sure to practice that occasionally is all I'm saying.

 

[oc_in_fw]

Cool. Just saying test restores periodically and document it. That's a nice CYA so if a necessary restore doesn't work, you can point at the fact that you've done the very same methodology and it worked. The whole reason for the back up is that it may be needed to recover from so be sure to practice that occasionally is all I'm saying.

We have 24 sims with at least two PCs each, and each PC has at least a two disk RAID. We have lost a lot of drives, and ShadowProtect always comes through.

 

[Chandu]

We have 24 sims with at least two PCs each, and each PC has at least a two disk RAID. We have lost a lot of drives, and ShadowProtect always comes through.

RAID won't always save you. I don't think you are doing anything wrong set up wise, but I would do real restores, say quarterly. The most unlikely of scenarios can hamper a restore.

 

[oc_in_fw]

The NAS’s are RAID6- and each sim has one. You can lose 2 drives on a NAS and still recover. I had that happen. I had one drive on a NAS disappear, and 2 hours into a rebuild another one crashed. Never saw that happen before, got I got it all fixed.

 

[StillShaving]

I think @Chandu is also warning about the larger problem of fire, theft, physical attack, or ransomware infection. There should always be some data backup offsite and periodic testing of the restoration.

 

[RayClem]

I have multiple computers in the house along with multiple cell phones, tablets, etc. For all of my important files, I use a Dropbox account to maintain backups. Any file placed in my Dropbox folder will be automatically synched to the Dropbox folders on my other computers as well as to the Dropbox servers in the cloud. Thus, if a computer goes bizzerk, a hard drive crashes, etc. I can recover my files. Even if my house is destroyed along with all my computers, I can still recover the files from the cloud. Is it 100% safe? I doubt anything in life is that certain other than eventual death, but it seems to be about as good as it gets.

I do not like relying solely on cloud storage as you are relying totally on an outside company to preserve your files. I like Dropbox because it backup files locally as well as to the cloud.

 

[Chandu]

The NAS’s are RAID6- and each sim has one. You can lose 2 drives on a NAS and still recover. I had that happen. I had one drive on a NAS disappear, and 2 hours into a rebuild another one crashed. Never saw that happen before, got I got it all fixed.

I've been out of the hardware side of servers for awhile, but it used to be that a controller could go bad so you write garbage to your entire raid array. No amount of drives are going to bail you out of that.

I think @Chandu is also warning about the larger problem of fire, theft, physical attack, or ransomware infection. There should always be some data backup offsite and periodic testing of the restoration.

There should always be an offsite backup. IronMountain or similar service is the best way. I'm also concerned that in IT the best laid plans of mice and men...

I say practice restores at least quarterly so that some software incompatibility or firmware bug or unknown corruption type thing doesn't rear it's head. Plus, if that is not the policy now, lobby for it and look good in the eyes of your management. Win Win.

RAID won't help you if you get a flaky disk controller that writes garbage. Though I do assume the verification has improved some over the years. In the old Seagate backup exec, I've seen stuff pass verification that was worthless to restore from.

 

[oc_in_fw]

I've been out of the hardware side of servers for awhile, but it used to be that a controller could go bad so you write garbage to your entire raid array. No amount of drives are going to bail you out of that.

There should always be an offsite backup. IronMountain or similar service is the best way. I'm also concerned that in IT the best laid plans of mice and men...

I say practice restores at least quarterly so that some software incompatibility or firmware bug or unknown corruption type thing doesn't rear it's head. Plus, if that is not the policy now, lobby for it and look good in the eyes of your management. Win Win.

RAID won't help you if you get a flaky disk controller that writes garbage. Though I do assume the verification has improved some over the years. In the old Seagate backup exec, I've seen stuff pass verification that was worthless to restore from.

We have had controllers go bad in PC. Only thing you can do is replace the controller and restore from the NAS. We haven’t had one go in a NAS yet (we will eventually, I am sure). The chance or losing a NAS and PC at the same time high, but if it were to happen I have got backups on external drives in a fireproof cabinet that will be less than one month old. The chance or losing all three at the same time is remote enough that management doesn’t want to add any more layers- backups already eat up enough of my time, I still have to work on simulators.

 

[Mick]

What brand NAS is that? We use Synology here at work. I have fallen into the role of Backup King here (as well as being Senior Lead Tech). I manage the backup jobs on 17 full flight simulators, each with at least an 8 computer network. The jobs are automatic (full backups on 1st and 15th, incrementals on M W F, we use ShadowProtect), but once a month I have to verify each and every backup- each verification takes at least 20 minutes. Of course, you can run more than one job at a time. On the 2nd of every month another backup occurs to a different folder on the NAS. These are then moved (by me) to external hard drives that are kept in a fireproof safe. May be a bit of overkill, but it only takes one disaster to make one rethink that opinion. I reckon I spend over 30 hours a month dealing with backup related activities. I am the only one willing to do it, so that helps to add a little job security.

Super micro 1U twins (two servers in a 1U chassis). I've been running this configuration exclusively since starting in business (2001).

In all that time 2 power supply failures, and a number of drive failures. I keep redundant/shelf spairs at the datacenter so P/S down time is a matter of minutes and drives are all RAID arrays so no down time.

I have 4 twins in service (8 servers).