【旧版・説明欄参照ください】サーバーレスアプリケーション向きの DB 設計ベストプラクティス

B
2019.05.09
Amazon Web Services Japan K.K.
Akihiro Tsukada, Startup Solutions Architect, Manager

9 B A 9 t D S p z O
• 9 B A 9 oml a
•
• 9 B A 9 e dg W O
k s n 9 B 0 1 ci[
• - 1 9 .
9 B A 9 2 1 2 p y
• A 9 ] h [| 9 1 1
• A 1 v PL A 9 rsb] 4

AWS Lambda
Amazon RDSAmazon Aurora
Amazon DynamoDB
Amazon ElastiCache Amazon Neptune
...etc
8

., 9 b
AWS Lambda
., @ a D B R
L ., 5
01 2
Amazon
RDS
Amazon
Aurora
9

a S
AWS Lambda
Amazon
RDS
Amazon
Aurora
D L
b L M 1
1 B 1 d
B R B . .
…
!
10

Private subnet
. B a IbD i
L P DN V R S
E D 2 C I
. L P ICE . C
e Id I
AWS Lambda Amazon
Aurora
Amazon
RDS
VPC
✕
Any Subnet
Elastic network interface
11

Private subnet
.1 0. aBb
RN LI L B L P
L PV 2 B C
.1 0. D d
E
AWS Lambda Amazon
Aurora
Amazon
RDS
VPC
Any Subnet
Elastic network interface
12

DuoB
• n D V S m pt uy
• D ACPS
m g m Oz B D
• elc am -
e U id
Amazon
DynamoDBAWS Lambda
…
!
14

B u eK 1 t m a
• r P
• r
a a B K
S D
• o m
• r A Q P
• 1 P ) (
) 1 ( )( IP
)( + ( +( :
K D a
• + 1 IP )
• i a P
r c
r
• r a
I D
https://docs.aws.amazon.com/ja_jp/amazondynamodb/latest/developerguide/best-practices.html
ao aP
n y m B
16

B o R O
Q • S M O B
tf R Q R
• l e
• tfR Q R B i
n a
• pf o R Q a
L
• L D c
O
E
• R 17

B c
IET • B Ic
• o E t
B t
R
• Re
I t
• m I
• I ED
B aI
• B I
• y
I
• n I
• B
18

B c
IET • B Ic
• o E t
B t
R
• Re
I t
• m I
• I ED
B aI
• B I
• y
I
• n I
• B
20

k E
(from re:Invent 2018 session)
•
c
• . 1
i E
• B e d
b R A a
• b d
21

B c
IET • B Ic
• o E t
B t
R
• Re
I t
• m I
• I ED
B aI
• B I
• y
I
• n I
• B
22

F IMKDP R E DI DI I
Use Case
C
C 2 A2 2
#
B
C
A2 2A A2 2A
A2 2A A2 2A
A2 2A 2 A2 2A A.
A2 2A 2 A2 2A
# 2A 2A C
2A 2 C
A2 A2
23

B c
IET • B Ic
• o E t
B t
R
• Re
I t
• m I
• I ED
B aI
• B I
• y
I
• n I
• B
24

.TABLE
Primary Key Attributes
PK SK, GSI-3-PK GSI-1-PK GSI-2-PK
Acr TargetACR AcrInfo1 AcrInfo2 EbookAsin
ABOOKACR1
ABOOKACR1-v0 ABOOK-ASIN1 ABOOK-SKU1
MAP-EBOOKACR1 SyncFileAcr ABOOK-ASIN1
ABOOKACR1#TRACK#1 ABOOK-ASIN1 ABOOK-SKU1
ABOOKACR1#TRACK#2 ABOOK-ASIN1 ABOOK-SKU1
EBOOKACR1 EBOOKACR1 EBOOK-SKU1 ASIN
25

.GSI1
GSI Partition Key Projected Attributes
GSI-1-PK Acr (Primary Table PK) TargetACR (Primary Table SK)
ABOOK-ASIN1 ABOOKACR1
ABOOKACR1-v1
ABOOKACR1#TRACK#1
ABOOKACR1#TRACK#2
SyncFileAcr ABOOKACR1 MAP-EBOOKACR1
EBOOK-SKU1 ABOOKACR1 EBOOKACR1
GSI2
GSI-2-PK Acr (Primary Table PK) TargetACR (Primary Table SK)
ABOOK-ASIN1 ABOOKACR1 MAP-EBOOKACR1
ABOOK-SKU1 ABOOKACR1
ABOOKACR1-v0
ABOOKACR1#TRACK#1
ABOOKACR1#TRACK#2
GSI3
TargetACR (Primary Table SK) Acr (Primary Table PK) TargetACR (Primary Table SK)
ABOOKACR1-v0 ABOOKACR1 ABOOKACR1-v0
MAP-EBOOKACR1 ABOOKACR1 MAP-EBOOKACR1
ABOOKACR1#TRACK#1 ABOOKACR1 ABOOKACR1#TRACK#1
ABOOKACR1#TRACK#2 ABOOKACR1 ABOOKACR1#TRACK#2
EBOOKACR1 EBOOKACR1 EBOOKACR1
26

B c
IET • B Ic
• o E t
B t
R
• Re
I t
• m I
• I ED
B aI
• B I
• y
I
• n I
• B
27

# Entity Use Case Lookup parameters Table / INDEX Key Conditions Filter Conditions
1 CompanionMapping getCompanionMappingsByAsin audiobookAsin/ebookSku GSI2 GSI-2=ABOOK-ASIN1 None
2 CompanionMapping
getCompanionMappingsByEbookAn
dAudiobookContentId
ebookAcr/sku,version,format or
audiobookAcr/asin,version,format
GSI-3 on TargetACR attribute OR
PrimaryKey on Table
GSI-3=MAP-EBOOKACR1 version=v and format=f
3 CompanionMapping getCompanionMappingsFromCache
ebookAcr/sku,version,format or
audiobookAcr/asin,version,format
GSI-3 on TargetACR attribute OR
PrimaryKey on Table
GSI-3=MAP-EBOOKACR1 version=v and format=f
4 CompanionMapping getCompanionMappings
syncfileAcr, ebookAcr?,
audiobookAcr?
GSI1 GSI-1=SyncFileAcr None
5 CompanionMapping getCompanionMappingsAvailable ebookAcr, audiobookAcr Primary Key on Table
Acr=ABOOKACR1 and
TargetACR beginswith "MAP-"
6 AcrInfo getACRInfo acr Primary Key on Table
Acr=ABOOKACR1 and
TargetACR beginswith
"ABOOKACR1-v"
7 AcrInfo getACRs acr / asin,version,format Primary Key on Table Acr=ABOOKACR1 version=v and format=f
8 AcrInfo getACRInfos acr Primary Key on table
Acr=ABOOKACR1 and
TargetACR beginswith "ABOOKACR1"
9 AcrInfo getACRInfosBySKU sku GSI2 GSI-2=ABOOK-SKU1
10 AudioProduct getAudioProductsForACRs acr Primary Key on table
Acr=ABOOKACR1 and TargetACR
beginswith "ABOOKACR1"
11 AudioProduct getAudioProducts sku, version, format GSI2 GSI-2=ABOOK-SKU1 version=v and format=f
12 AudioProduct
deleteAudioProductsMatchingSkuV
ersions
sku, version GSI2 GSI-2=ABOOK-SKU1 version=v
13 AudioProduct getChildAudioProductsForSKU sku GSI2 GSI-2=ABOOK-SKU1
14 Product getProductInfoByAsins asin GSI1 GSI-1=ABOOK-ASIN1
15 Product getParentChildDataByParentAsins asin GSI1 GSI-1=ABOOK-ASIN1
16 AudioFile getAudioFilesForACR acr Primary Key on table
Acr=ABOOKACR1 and
TargetACR beginswith
"ABOOKACR1#"
28

D e m a
IET • mD e m
• I
D E
R
• y a R m yB
m I
• ot I B
• R I
m D c I
• D I
• R I
m B n
• m I
• D I
29

E
ID:
:
AWS Loft Tokyo:
:
:
ID
2 DynamoDB
AWS Loft
Tokyo
yy/3/4
#DynamoDB
#Serverless
14
AWS Loft
Tokyo
yy/5/9
#Serverless
#Lambda
#Design
33 API Night
AWS Loft
Tokyo
yy/6/23
#WebAPI #JWT
#APIGateway
#Auth #OAuth
#OIDC #Cognito
> >><< <
33

D e m a
IET • mD e m
• I
D E
R
• y a R m yB
m I
• ot I B
• R I
m D c I
• D I
• R I
m B n
• m I
• D I
34

id (PK)
name
venue_id (FK)
date
Events
id (PK)
name
Tags
event_id (PK, FK)
tag_id (PK, FK)
EventTags
id (PK)
name
address
Venues
-- SQL
-- :
SELECT e.id, v.name... FROM Events e
JOIN Venues v ON v.id = e.venue_id
JOIN EventTags et ON et.event_id = e.id
JOIN Tags t ON t.id = et.tag_id
WHERE v.name = :vname AND t.name = :tname
35

2 2 2
Use Case
. . . .
. . . .
. . .
. . . .
# . . .
. .
. .
36

i n t hc et Ql tTABLE
Partition Key Attributes
{EventID} {EventName} {VenueName} {TagName1, ...} {yyyy-mm-dd}
... ... ... ... ...
• . . . .3 GB I S
G D +- 3 . Q G
• +- 3 i n t hc m t
lma t Sr c
• QE +- 3 Q F D vQ D
u +- 3 QE . . . -/ . . S
D
BS y F
38

e b l cTABLE
Partition Key GSI-1-PK GSI-2-PK GSI-3-PK GSI-4-PK
{EventID} {EventName} {VenueName} {TagName1, ...} {yyyy-mm-dd}
... ... ... ... ...
• / + / . 0- 2 0-. 3.2
G
• S u b le b r a b b
• l o y x
b dn c Q I
39

GTABLE
PK, GSI-1-SK SK GSI-1-PK GSI-2-PK
ID DataType DataValue VenueName VenueAddress
{EventID} EventName {EventName}
{EventID} VenueID {VenueID}
{EventID} Date {yyyy-mm-dd}
{EventID} Tag_{TagName} Tag_{TagName}
{VenueID} VenueInfo {VenueName} {Address}
• S 3
• . I G
• 3 1
41

3 3GSI1
GSI-1-PK GSI-1-SK (Primary Table SK)
DataValue ID DataType VenueAddress
{EventName} {EventID} EventName
{VenueID} {EventID} Venue
{yyyy-mm-dd} {EventID} Date
Tag_{TagName} {EventID} Tag_{TagName}
GSI2
GSI-2-PK (Primary Table PK)
VenueName ID VenueAddress
{VenueName} {VenueID} {Address}
• . . 42

TABLE
E123 EventName DynamoDB
E123 VenueID V32
E123 Date yy/3/4
E123 Tag_#DynamoDB Tag_#DynamoDB
E123 Tag_#Serverless Tag_#Serverless
E145 EventName
E145 VenueID V32
E145 Date yy/5/9
E145 Tag_#Lambda Tag_#Lambda
E145 Tag_#Design Tag_#Design
V32 VenueInfo AWS Loft Tokyo
• . .4
• . 4 2 E
51 4 3
43

TABLE
E123 VenueID V32
E123 Date yy/3/4
E145 EventName
E145 VenueID V32
E145 Date yy/5/9
1 3
Query(ID = :eventId) . 3 : .
Table Query(ID = "E123") 44

TABLE
E123 VenueID V32
E123 Date yy/3/4
E145 EventName
E145 VenueID V32
E145 Date yy/5/9
1 3
Query(DataValue = :tagName) . : .
GSI1.Query(DataValue = "Tag_#Serverless") 45

dcj nlr
Use Case Parameters Table/Index API & Key Conditions
. DI I. DI . DI1- G C G B 4 G 1- ) ( DI1
. DI I. DI . DI2 C 1 4 G - I B ) ( DI2 C
. DI I. DI D 2 C
1
1
4 G D 2 C ) ( D 2 C
4 G - I B ) ( D 1
. DI I. DI - I 1 4 G - I B ) ( I
# . DI I. DI 1 4 G - I B ) (I 2 C
I . DI1- G C G B 4 G 1- ) ( DI1
D I D . DI1-
G C G B
G C G B
I1I C 1- ) ( DI1 D - I F ) D 1-
I1I C 1- ) ( D 1 D - I F ) D 1D=E
• i fe f B 1D bp N Sh gb
m aV TP I1I C I I:?1I C 4 G : D Qbus
• / BI G ,ED I ED GE :I ED . FG ED bot
46

A
• . . . a
. 1 B4
• y 2 D
•
n 3 .
. .
. o
o
• m z
o 3 48

s ztL P
• 9 B A 9 rpo [d
• A 9 h cg ]a
• 9 B A 9 h cg ]a
n 9 B 0 1 l
• - 1 9 .
9 B A 9 2 1 2 s
• A 9 bD ibkm 9 1 1
• A 1 y S OW A 9 v e Dm 49

D GI1 E A S D GI
TABLE
PK, GSI-1-SK SK GSI-1-PK
ID DataType DataValue1 DataValue2
{EventID} VenueInfo {VenueName} {Address}
{EventID} Tag_{TagName} Tag_{TagName}
•
. - D
52

-1 .TABLE
PK, GSI-1-SK SK GSI-1-PK
ID DataType DataValue1 DataValue2
E123 VenueInfo AWS Loft Tokyo
E123 Date yy/3/4
E145 EventName
E145 VenueID AWS Loft Tokyo
E145 Date yy/5/9
53

1 -. 1 -.GSI1
DataValue ID DataValue1 DataValue2
{EventName} {EventID} EventName
{VenueID} {EventID} Venue
{yyyy-mm-dd} {EventID} Date
Tag_{TagName} {EventID} Tag_{TagName}
{VenueName} {EventID} {VenueName} {Address}
54

(BB =I _iurmv
Use Case Parameters Table/Index API & Key Conditions
G ED E G E) G E. C= C 1 C . G E.
G ED E G ED) G E/ - . 1 C E G E/
G ED E G ED) / - . 1 C E G /
G ED E G ED) E - . 1 C E E
# G ED E G ED) - . 1 C E E /
D E D) G E. C= C
1 C
. G E. E B = D =E
E ) G E. C= C
- E.E
. G E E B . :A
• T o w h a i e nr V
• tl fd y xP CAD A D y Q NpS sg b 55

2 U m E SVg I d
1 -
1 - .
d nm
A- . B
d , .
e
x um
• V I E EU G E U
• G 2 d I
a S a m d
• i s
• A my t e VB
• ir v p NS S G E
56

- 2.
TABLE
PK SK, GSI-1-PK GSI-1-SK
EventID DataType DataValue
{EventID} VenueName {VenueName}
{EventID} VenueAddress {VenueAddress}
{EventID} Tag_{n} {TagName}
58

2 - .
TABLE
PK SK, GSI-1-PK GSI-1-SK
EventID DataType DataValue
E123 VenueName AWS Loft Tokyo
E123 VenueAddress
E123 Date yy/3/4
E123 Tag_1 #DynamoDB
E123 Tag_2 #Serverless
E145 EventName
E145 VenueName AWS Loft Tokyo
E145 VenueAddress
E145 Date yy/5/9
E145 Tag_1 #Serverless
E145 Tag_2 #Lambda
E145 Tag_3 #Design
59

2 . -GSI1
DataType DataValue EventID
EventName {EventName} {EventID}
VenueName {VenueName} {EventID}
VenueAddress {VenueAddress} {EventID}
Date {yyyy-mm-dd} {EventID}
Tag_{n} {TagName} {EventID}
60

:: TS
Use Case Parameters Table/Index Key Conditions
= = = A = - 1 A 12 = =
= = = A = 1 .
1=1 A: = 1
= 1 = 1
= = = A 1 .
1=1 A: 1
1 1
# = = = A 1= .
1=1 A: 1=
1= 1=
= = = A 1
4825
6
57 3 1
1 = 1 A 1 1
1 = 1 A =
= A =
• P = = A 1 I N P BDaG V E
61

. - N . . N PB
GSI1
DataType DataValue EventID
EventName {EventName} {EventID}
VenueName {VenueName} {EventID}
VenueAddress {VenueAddress} {EventID}
Date {yyyy-mm-dd} {EventID}
Tag_{n} {TagName} {EventID}
• 2. BS . b KT
Query(DataType = begins_with("Tag_") and DataValue = :tagName)
. _ . B AN
• Query(DataType = Tag_{n} and DataValue = :tagName)
dBS Pe TE
. a N
62

【旧版・説明欄参照ください】サーバーレスアプリケーション向きの DB 設計ベストプラクティス

Recommended

Recommended

More Related Content

More from Amazon Web Services Japan

More from Amazon Web Services Japan (20)

Recently uploaded

Recently uploaded (20)