M16-WAFLSimplified

7/27/2019 M16-WAFLSimplified

1/21


2/21

2009 NetApp. All rights reserved.

Module Objectives

By the end of this module, you should be able to:

Describe how data is written to and read from

a WAFL file system on a volume

Describe the WAFL file system, including

consistency points, RAID management, andstorage levels

Describe how RAID is used to protect disk data

Describe how the WAFL file system processeswrite and read requests


3/21


Data ONTAP 8.0 7-Mode Architecture

StorageRAIDProtocolsNetwork

D-Blade

NVRAM

WAFL

Physical

Memory

FreeBSD

Disk

Array

Clients

M-

Host

Clien

tProtocol

A

ccess


4/21


5/21


Write Requests

Write requests are received by Data ONTAP

through multiple protocols:

CIFS

NFS

FCP

iSCSI

HTTP

WebDAV Write requests are buffered into:

System memory

NVRAM


6/21


Write Request Data Flow: Write Buffer

Network

NetworkStack

RS-232

HBA

NIC

Protocols

SAN

Service

NFS

Service

CIFS

Service

Memory Buffer

NV

R

A

M

NVRAM Full

WAFLSAN Host

UNIX

Client

WindowsClient

NVLOGNVLOG

NVLOG

NVLOG

NVLOG

Storage

RAID


7/21 2009 NetApp. All rights reserved.

Consistency Point

A consistency point (CP) is a completely self-

consistent image of the file system. Although

dynamic, if the file system were frozen

momentarily to capture its structurethat is a

CPA CP occurs when designated data is written

to a disk and a new root inode is determined

A CP occurs for multiple reasons, including the

following:

One bank of the NVRAM card is full

10 seconds have elapsed



Consistency Point (Cont.)

During a CP, Data ONTAP flushes writes to

disk

Always writes to new data blocks

The volume is always consistent on disk

When Data ONTAP flushes memory to disk: It updates the file system atomically, meaning

that the entire write must be completed or the

entire write is rolled back

This includes all metadata

After checking, the NVRAM is cleared



WAFL

Write Request Data Flow: WAFL to RAID

Network

NetworkStack

RS-232

HBA

NIC

Protocols

SAN

Service

NFS

Service

CIFS

Service

Memory Buffer

Storage

NV

R

A

M

NVLOGNVLOG

NVLOG

NVLOG

NVLOG

NVRAM Full

RAID

SAN Host

UNIX

Client

WindowsClient



Consistency Point: WAFL to RAID

The RAID layer calculates the parity of thedata:

To protect it from one or more disk failures

To protect stripes of data

Also calculates block or zone checksums If a data disk fails, the missing information can

be calculated from parity

The storage system can be configured as

either: RAID 4Allows one disk failure in the RAID

group

RAID-DPAllows up to two disk failures in

the RAID group



Write Request Data Flow: RAID to Storage

Network

NetworkStack

RS-232

HBA

NIC

Protocols

SAN

Service

NFS

Service

CIFS

Service

Memory Buffer

RAID

Storage

NV

R

A

M

NVLOGNVLOG

NVLOG

NVLOG

NVLOG

NVRAM Full

WAFL

4k

SAN Host

UNIX

Client

WindowsClient Block or zone

checksum

computed



Consistency Point: RAID to Storage

Storage layer commits the data and parity to

the physical disks

The root inode is updated to point to the new

file inodes on the disk

The NVRAM is flushed and made available The CP now is complete



Storage

Write Request Data Flow: Storage Writes

Network

NetworkStack

RS-232

HBA

NIC

Protocols

SAN

Service

NFS

Service

CIFS

Service

Memory Buffer

NV

R

A

M

NVLOGNVLOG

NVLOG

NVLOG

NVLOG

NVRAM Full

WAFL

SAN Host

UNIX

Client

WindowsClient

RAID



NVRAM

Data ONTAP writes from system memory

NVRAM is never used for normal write

operations

NVRAM is backed up with a battery

If a system failure occurs before thecompletion of a consistency point, the data is

read from NVRAM and added back to system

memory buffer when the system is brought

back online (or by the partner machine in ahigh-availability controller configuration)



Read Requests



Read Requests

Every time a read request is received, WAFL

does one of the following:

Reads the data from the system memory, also

known as cache

Reads the data from the disks The cache is populated by:

Data recently written to disk

Data recently read from disk



Read Request Data Flow: Cache

SAN Host

UNIX

Client

WindowsClient

RS-232

HBA

NIC

SAN

Service

NFS

Service

CIFS

Service

Memory Buffer

N

V

R

A

M

Cache

Network

NetworkStack Protocols

WAFL

RAID

Storage



Read Request Data Flow: Read from Disk

SAN Host

UNIX

Client

WindowsClient

RS-232

HBA

NIC

SAN

Service

NFS

Service

CIFS

Service

Memory Buffer

N

V

R

A

M

WAFL

Cache

Storage

Network

NetworkStack Protocols

RAID



Module Summary

In this module, you should have learned to:

Describe how data is written to and read from

a WAFL file system on a volume

Describe the WAFL file system, including

consistency points, RAID management, andstorage levels

Describe how RAID is used to protect disk data

Describe how the WAFL file system processeswrite and read requests


20/21

Exercise

Module 16: WAFL Simplified

Estimated Time: 15 minutes


21/21

Check Your Understanding

What is a consistency point?

A completely self-consistent image of the entire

file system

What is the purpose of RAID?

To protect dataRAID is the method used toprotect disk data by which a parity value is

calculated across the disks in 4-KB blocks

What is the storage layer?

Handles the I/Os to disk and checks for disk

failures

M16-WAFLSimplified

Documents

Transcript of M16-WAFLSimplified