now that we've discussed incident response and forensic investigations, the next four to security operations that were concerned with his fault management. So we'll talk about the various ways that we guarantee availability on our network,
whether its availability of hard drives or servers themselves or even
personnel. So when we do talk about redundancy, one of the things we think about is having spares around. You can have hot spares that are already installed and functioning. It's just a matter of switching over to them or colds fairs. They're downstairs in the closet.
But the bottom line is we want some sort of redundancy of devices
now. How often What's the life span of a device? Do I need a spare dough? I need more than one. Well, a lot of times, what we'll do is we'll look to a metric cult in TPF. Meantime, between failures
and that's the life span of the device. So if I look at a hard drive and it has an empty B F of three years, that's how long I expect it to last.
Also, keep in mind that for redundancy sake and fault tolerance, service level agreements can come in very handy because that's the commitment from our vendors. Thio. Help us keep our up time going
now. The next feature that we would look at for redundancy is rate redundant array of inexpensive disks or inexpensive devices or independent devices. Honestly, it really depends on what the last book you read calls it because
they're different names. But ultimately the important pieces redundant array.
Now what we're looking for here is multiple hard drives. Acting is a single logical unit for the purpose of fault tolerance. So the only rates I really want you to be concerned with for the exam 01 and five. Well, if we start off with Raid Zero,
we're gonna find that's not redundant at all. There is no fault tolerance
with Raid zero that's referred to his disk striping.
And with this striping, let's say I have to physical discs. Data is striped across disc one and two simultaneously. So, like, for instance, if I have a 24 k file
ah, the 1st 12 k is being written two disc one and at the same time, the 2nd 12 K's being written the disc to. So we say that that data striped across multiple disks.
But the problem with that is, if a disc fails, you have the potential to have lost all your data.
All right, so I even use it speed performance. And that's it's big benefit. It has benefits with both read and write.
Now, if we do want thought, fault, tolerance, which we probably do, we're gonna look perhaps to raid one raid one is disc mirroring. And with dis mirroring essentially, what we have is to replicants. They are exactly the same. They're exactly identical.
So the two discs are, um,
as I mentioned replicas of each other. And if one fails, it's very easy to switch over now, do you keep in mind you're always gonna lose 50% of your disk space to fault tolerance their exact copies. I got 25 gigabyte drives. That's 10 gigabytes, really should say terabytes. But
attn any right? I'm gonna lose one of those drives
for the purpose of fault. Howard's now Raid five is disc striping with parody. So it gives me the speed boost that describing does. But the parody is striked out across multiple disks and can be used to rebuild the information. Should we lose a drive? All right, so we have rate zero