Burned by Eschewing Best Practices
-
@Romo said:
I use mdadm with our kvm vm hosts, I haven't used xenserver but I would think mdadm would also be a possibility for him
Yes, works the same. But for someone who is new to Linux this will be a shock to his system.
-
This guy is in for a tough time. No AD for 40 machines..
http://community.spiceworks.com/topic/1274733-40-workgroup-pc-s
-
@scottalanmiller said:
This guy is in for a tough time. No AD for 40 machines..
http://community.spiceworks.com/topic/1274733-40-workgroup-pc-s
...and all are home versions, some Vista.... ewwwww
-
Yeah, it is going to be tough. No AD with forty machines. Not fun. But I've seen way bigger without AD, it's not that bad. I'd rather AD, of course. But you can do a lot without it.
-
That's a lot of PCs to have no Active Directory. Going to need to do ridiculous things like having common admin passwords or using Keepass and storing tons of passwords.
-
Not a bad approach.
-
Here is another topic of it, where an IT Department virtualized, but failed to develop and deploy a proper backup solution. Now the IT person is stuck manually restoring files and ActiveDirectory from a Snapshot created in January!
-
@DustinB3403 I wonder if that snapshot from January is part of what messed his system up when it lost power. You shouldn't keep snapshots hanging around that long. Nothing but trouble.
-
Well the story appears to thicken, as he was panicked (not having a recovery strategy) and simply restored with what he saw...
So likely it did.
-
@DustinB3403 Panicking leads to bad choices.
-
@brianlittlejohn said:
@DustinB3403 Panicking leads to bad choices.
And bad choices lead to panicking.
-
And the plot thickens even more.
he has separate DC's, that are for an Old DC.
So only having 1 DC. More of that eschewing practice...
-
Here is another, RAID 5 (4x3TB) took the server down to replace the drive. (May have been the only way to do so). Restored AD from backup etc etc.
-
@DustinB3403 said:
So only having 1 DC. More of that eschewing practice...
Sometimes it is OK to do that, as I learned here at ML.
Risk and reward. Reliability not redundancy.
http://www.mangolassi.it/topic/6495/storage-question/37 -
@BRRABill said:
@DustinB3403 said:
So only having 1 DC. More of that eschewing practice...
Sometimes it is OK to do that, as I learned here at ML.
Risk and reward. Reliability not redundancy.
http://www.mangolassi.it/topic/6495/storage-question/37But here the risk is losing all ability to use your computer systems. So the risk does not outweigh the cost.
-
@DustinB3403 said:
But here the risk is losing all ability to use your computer systems. So the risk does not outweigh the cost.
By losing the DC?
-
In that case yes, a single DC which was restored from a backup make in January.
The system is almost useless. A second VM host, even a desktop running HyperV and a vDC on it would have prevented his predicament.
-
The thread I posted made a good case that it isn't crazy for smaller businesses to run only 1 DC, even though it is not "best practice".
@scottalanmiller said that for small businesses, it is actually decently rare that the cost of a second AD DC is justified.
Not saying it is optimal, and of course there is risk. But there was a lot of talk in that thread about only having 1 DC.
-
But @BRRABill what you're missing is that the OP has multiple issues, compounding the issues, no recovery solution, no documentation, nothing.
All of this compounds the issue.
A very simple answer to that is a super cheap backup Hypervisor running an "Oh-Shit Domain controller".
-
Right, there are other issues.
Just saying having 1 DC isn't what caused the problem, and it the right circumstances might have been fine.
I'm just defending the 1 DC theory.
I agree all the rest of the stuff is nowhere near best practice.
Like if you don't want to do backups, perhaps have a second DC DEFINITELY.
