RAID and Data Backups

Video Activity
Join over 3 million cybersecurity professionals advancing their career
Sign up with
Required fields are marked with an *
or

Already have an account? Sign In »

Time
7 hours 50 minutes
Difficulty
Beginner
CEU/CPE
8
Video Transcription
00:00
>> When it comes to providing redundancy for hard drives,
00:00
the first thing you should think of is RAID,
00:00
Redundant Array of Independent Devices,
00:00
or you could hear independent disks.
00:00
That's fine either way. The idea
00:00
is there's redundant array.
00:00
We have multiple disks acting as a single logical disk,
00:00
and it's usually for the purpose of redundancy.
00:00
That being said, the very first RAID we look at,
00:00
RAID 0 is not redundant.
00:00
If I rule the world, I would call it AID.
00:00
It's an array of independent disks,
00:00
but it's not redundant.
00:00
Without redundancy, what that
00:00
means is if one of your disks fails,
00:00
you run the possibility of losing all your data.
00:00
What RAID 0 does is it takes
00:00
two physical disks and
00:00
uses a feature called disk striping.
00:00
If there's 24 kilobytes worth
00:00
of data to be written to the disk,
00:00
12 kilobytes gets written to disk 1,
00:00
and simultaneously,
00:00
the other 12 kilobytes is written to disk 2.
00:00
You get a simultaneous RAID
00:00
>> instead of a sequential RAID.
00:00
>> It saves time, it's a speed improvement.
00:00
Not only is it faster for writing,
00:00
but it's also faster for reading because its data is
00:00
being located on drive 1 or is being read on drive 1,
00:00
it's being located on drive 2.
00:00
You get a performance boost,
00:00
but because of the fact that it's not
00:00
redundant is not really a
00:00
>> great choice for most servers.
00:00
>> This would be good, however,
00:00
on media [inaudible] servers
00:00
where we get that performance,
00:00
but we don't really have the need for fault tolerance.
00:00
If high availability is my need, RAID 1.
00:00
What RAID 1 is,
00:00
is it's disk mirroring.
00:00
I have two disks, each one
00:00
>> an exact replica of the other.
00:00
>> This provides really high availability
00:00
because if there is a failure of one of the disks,
00:00
it's very easy and very
00:00
quick to transfer over to the other.
00:00
You hardly have any downtime at all.
00:00
Drawback there is, is I'll
00:00
always lose half of my disk space.
00:00
By that, I mean if I go out and
00:00
buy two five-terabyte drives,
00:00
I have 10 terabytes of space.
00:00
Five terabytes is for redundancy,
00:00
so I don't get to use that per se.
00:00
That's one of the drawbacks to disk mirroring,
00:00
but really good redundancy.
00:00
RAID 5, disk striping with parity.
00:00
This takes a minimum of three physical disks.
00:00
I have disk striping,
00:00
which gives me the speed improvement.
00:00
But also on each disk contains parity for another disk.
00:00
Ultimately, if disk 1 fails,
00:00
parity on disk 2 can be
00:00
used to rebuild the disk that's failed.
00:00
That being said, this is
00:00
a dicey process and sometimes rebuilding the disk
00:00
can actually cause data loss or
00:00
a disk failure. It's not ideal.
00:00
RAID 6 is coming around to replace RAID 5.
00:00
Here, we need four physical disks.
00:00
By using four physical disks,
00:00
two disks are for fault tolerance,
00:00
so not as many as two drives can fail.
00:00
That's a little bit better of the redundancy.
00:00
RAID 10 or sometimes referred to as 1 plus 0,
00:00
we said disk striping.
00:00
The stripes isn't fault-tolerant.
00:00
What if I mirror that stripes
00:00
>> at to another set of disks?
00:00
>> That's what RAID 10 is. That requires
00:00
four physical disks as well.
00:00
Of course, half of my space would
00:00
be used for fault tolerance.
00:00
RAID is always discussed on the exam.
00:00
Expect to get questions,
00:00
expect to get a couple of performance-based questions.
00:00
Thou shalt know thy RAID.
00:00
Another hot topic is going to be backups
00:00
because even though we have RAID in place,
00:00
file corruption malware
00:00
>> can infect the entire RAID array.
00:00
>> RAID doesn't really protect our data,
00:00
that way we need it to be
00:00
protected so we back up our data.
00:00
There are multiple types of backups we can use.
00:00
There is a full and incremental,
00:00
a differential, and a copy.
00:00
The full is the easiest to understand
00:00
because a full backup backs everything.
00:00
Incremental backup backs up what's
00:00
changed since the last backup of any kind.
00:00
We do a full backup on Sunday night,
00:00
and we do an incremental on Monday
00:00
that backs up what's changed since Sunday.
00:00
We do another incremental on Tuesday,
00:00
and backs up what's changed since Monday.
00:00
An incremental Wednesday backs up,
00:00
which changed since Tuesday.
00:00
This takes less time in backup.
00:00
However, in the event that we need to restore,
00:00
we have to restore the full backup
00:00
in each day's incremental.
00:00
That can take a little longer.
00:00
A differential backup backs up
00:00
which changed since the last full backup.
00:00
We do a full backup on Sunday.
00:00
Monday is incremental and
00:00
backs up which changed since Tuesday.
00:00
Tuesday is incremental, what's changed since Sunday.
00:00
Wednesday is incremental, which changed since Sunday.
00:00
This is going to give us a quicker
00:00
restore because we have to
00:00
restore the full backup
00:00
>> and the most recent differential.
00:00
>> All this has to do in
00:00
Windows environments with a little bit
00:00
called the archive bit.
00:00
The archive bit is just a flag that pops up and says,
00:00
I've changed. I need to be backed up.
00:00
When you do a full backup,
00:00
all the bits get reset.
00:00
That's just a way of acknowledging,
00:00
I've backed up everything.
00:00
Everything that needs to be backed up is backed up.
00:00
On Monday, as files start to change,
00:00
those flags pop up.
00:00
We do an incremental Monday night,
00:00
now backs up everything for the day and it
00:00
clears the bits and the signals it's taken care of.
00:00
Tuesday's new files are modified.
00:00
The file pops back up.
00:00
Tuesday's incremental backs up which
00:00
changed on Tuesday and clears the Bits.
00:00
The differential backup
00:00
>> is different from the other two,
00:00
>> because the differential Bit
00:00
backup does not clear the Bit,
00:00
which is exactly why Monday's files changed.
00:00
It backs up everything that's changed since
00:00
Sunday, the Bit isn't cleared.
00:00
As files pop up Tuesday,
00:00
you have the things for Monday with
00:00
flag set and you have the things from Tuesday.
00:00
We back up everything since the last full backup.
00:00
There's also a copy. With virtual machines,
00:00
you can copy specific files
00:00
or you could do a full backup.
00:00
This just does not reset the archive bit.
00:00
Also with virtualization, we now
00:00
think about just reverting to Snapshots.
00:00
Whatever our strategy is,
00:00
we need to make sure that we have
00:00
fault tolerance for our data
00:00
>> as well as our hard drives.
Up Next