Você está na página 1de 30

Module 2: Data Domain Operating Environment

Based on
Linux
Includes (but not limited to)

Data Domain
Enterprise Manager

EMC Education Services

licensed
features
DD
Boost
ddvar &
backup file
systems

CLI
DIA

NFS, CIFS,
NDMP, VTL

deduplication
SISL

Lesson 1: Data Domain Deduplication


Introduction

Data Domain deduplication stores only one copy of


data
In this lesson you learn more about Data Domain
deduplication

Objective

Describe Data Domain deduplication

EMC Education Services

How Data Domain Deduplication Works


file system /VTL backup /archive SW
data stream
1st full backup
A

2nd full backup

1st increment
D

E F

unique variable segments (4KB-12KB)


redundant data segments
compressed unique segments

EMC Education Services

Lesson 2: SISL
Introduction

Data Domain Stream-Informed Segment Layout (SISL)


is an important scaling architecture that helps give
Data Domain systems speed
In this lesson you learn more about SISL

Objective

Describe SISL

EMC Education Services

SISL Definition

Important Data Domain architecture

Technology pioneered by Data Domain

Provides fast & efficient deduplication

99% of duplicate data segments identified inline in RAM


before they are stored to disk
System throughput increases directly as CPU performance
increases
Minimizes disk footprint by minimizing disk access

EMC Education Services

Segment
Data sliced into segments
1Deduplication

Fingerprint

Filter

Segments given fingerprint ID (segment ID)


Fingerprint IDs compared to fingerprints in cache
1.
If fingerprint
ID new, continue
1
backup
2.
If fingerprint ID duplicate, reference, then delete

4 Compress

Groups of new segments compressed using common technique


(lz, gz, gzfast)

5 Write

Segments (including fingerprints, metadata, & logs)


written to containers, containers written to disk

~~~~~
~~~~~
~~~~~

~~~~~
~~~~~
~~~~~

~~~~~
~~~~~
~~~~~

~~~~~
~~~~~
~~~~~

~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~ ~~~~~
~~~~~ ~~~~~
~~~~~ ~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~
~~~~~

2
1

4
Data Domain system

EMC Education Services

container
disk

disk

Lesson 3: DIA
Introduction

Data invulnerability architecture (DIA) provides safe


& reliable storage

Objective

Describe DIA

EMC Education Services

DIA Definition

Provides safe & reliable storage

Fights data loss 4 ways


1.

End-to-end verification

2.

Fault avoidance & containment

3.

Continuous fault detection & healing

4.

File system recovery

EMC Education Services

DIA End-to-End Verification

EMC Education Services

DIA Fault Avoidance & Containment

EMC Education Services

10

DIA Continuous Fault Detection & Healing

EMC Education Services

11

DIA File System Recovery

EMC Education Services

12

Lesson 4: File Systems


Introduction
A Data Domain system includes 2 file systems
1.

Administrative (ddvar)

2.

Storage (Mtree)

Objectives
.

Describe the administration file system (ddvar)

Describe the storage file system (Mtree)

EMC Education Services

13

Administration File System

ddvar
NFS directory /ddvar

CIFS share \ddvar

Contains Data Domain system core & log files

Contains configuration information

Cant rename or delete

Cant access all sub-directories (e.g. core)

EMC Education Services

14

Storage File System

NFS directory /backup


data

CIFS subfolder \backup


Destination directory for deduplicated data
Comes pre-configured for NFS export as /backup
Clients performing backup & restore must access
/backup
You configure directory export levels to separate
& organize backup files

EMC Education Services

col1
backup
Mtree
MTree2

MTree3

15

Storage File System (Continued)

Uses compression

Implements data integrity

Reclaims storage space with file system cleaning

Identifies patterns in stored data

data
col1
backup
Mtree
MTree2

Supported maximum data streams change


per OS version & Data Domain model

EMC Education Services

MTree3

16

Storage File System (Continued)


You cant add anything

data

col1
backup
Mtree

You cant delete/rename


/a

Mtree

You can add up to 14


directory Mtrees under
/data/col1
You can manage each
Mtree directory separately
(for example, different
compression rates)

/b
Mtree

EMC Education Services

/a
/b

17

Lesson 5: Data Domain Data Paths


NFS
CIFS
DD Boost
NDMP
VTL

backup
server

Introduction

Data Domain data paths include


NFS, CIFS, DD Boost, NDMP & VTL

Data
Objectives
Includes
(but not limited to)
Domain
system

Indicate where Data Domain


Licensedsystem
Features

fits into typical backup


environment
CLI

Data Domain
Enterprise Manager

EMC Education Services

Deduplication
Identify
data path for NFS, CIFS,
DD Boost, NDMP over Ethernet

Identify data path for Data Domain


VTL over Fibre Channel
18

Data Domain System in Typical Backup


Environments
Solaris, Oracle, Linux,
Windows, SQL,
Exchange &
application servers
production LAN - Gig-E copper & fibre
backup server

tape library

backup server

copy to tape quarterly if needed

EMC Education Services

WAN-based replication
WAN
offsite disaster
recovery location

19

Data Path over Ethernet


backup/archive media servers

DD Boost
NFS/CIFS
FTP/NDMP

Ethernet

TCP(UDP)/IP

deduplicated replication

Ethernet
NFS/CIFS/DD Boost
FTP/NDMP

Data Domain system

EMC Education Services

TCP(UDP)/IP

WAN

Ethernet

deduplicated
data written
to /backup
file system
Data Domain system

20

Data Path over Fibre Channel VTL


backup/archive media servers

/dev/rmt
\\.\Tape#
WAN
SAN

FC SAN

TCP(UDP)/IP

deduplicated replication

Ethernet

Ethernet

VTL
deduplicated
data written to
backup
file system
Data Domain system

EMC Education Services

Data Domain system

21

Lesson 6: Administration Interfaces


Introduction
You can access a Data Domain system in 2 ways:
Enterprise Manager & CLI

Objectives
Describe administration interfaces

Use administration interfaces

EMC Education Services

22

Access Enterprise Manager


DD hostname
http://ddhostname/ddem

EMC Education Services

23

Enterprise Manager Main Screen


You need the sysadmin password
to add a Data Domain system.

monitored
systems

EMC Education Services

cumulative
information
for monitored
systems

24

Enterprise Manager Tabs

Select a system

sub-level
administration
tabs

EMC Education Services

top-level
administration
tabs
sub-level
tab
admin
interface
display

25

CLI

Access CLI via SSH, serial console, telnet, keyboard & monitor

keyboard

eth0a

video port
serial port

eth0b

EMC Education Services

26

CLI (Continued)

Type help to list main


commands

Type help & any


keyword to search
online help

See Command Reference Guide for complete details


EMC Education Services

27

Module Review

Data Domain deduplication is performed inline on


bytes, not files

SISL gives Data Domain systems speed

DIA provides safe & reliable storage

DIA fights data loss 4 ways


1.

End-to-end verification

2.

Fault avoidance & containment

3.

Continuous fault detection & healing

4.

File system recovery

EMC Education Services

28

Module Review (Continued)

ddvar contains system & core files & is used for


administration
Mtrees (including the backup directory) contain storage
data

The backup directory cant be moved/renamed

You can add/change Mtrees

Data Domain data paths include NFS, CIFS, DD Boost,


NDMP, & VTL

EMC Education Services

29

Module Review (Continued)

A Data Domain system has 2 administration


interfaces
1.

Enterprise Manager

2.

CLI

You can perform most operations in Enterprise


Manager

EMC Education Services

30