Big query - Command line tools and Tips - (MOSG)

Google BigQuery
- Command line and Tips -
2016/06/08
Mulodo Vietnam Co., Ltd.

What’s BigQuery
Official site : https://cloud.google.com/bigquery/docs/
BigQuery is Google's fully managed, petabyte scale, low
cost analytics data warehouse.
BigQuery is NoOps—there is no infrastructure to manage
and you don't need a database administrator—so you can
focus on analyzing data to find meaningful insights, use
familiar SQL, and take advantage of our pay-as-you-go
model.
→ DWH: SQL like (easy to use), Petabyte scale(for Huge data)

Previous study
“BigQuery - The First Step -“ (2016/05/26)
• Just try to start for Google Big Query
• Using query on the Google Cloud Platform console.
• Create your own Dataset and Table
• Using query for your table GPC console.
http://www.meetup.com/Open-Study-Group-Saigon/events/231233151/
http://www.slideshare.net/nemo-mulodo/big-query-the-first-step-mosg
c.f. “Big Data - Overview - “
http://www.slide http://www.meetup.com/Open-Study-Group-Saigon/events/229243903/
share.net/nemo-mulodo/big-data-overview-mosg

Command line tools and Tips
1. Preparation (install SDK and settings)
2. Try command line tools
create datasets, tables and insert data.
3. Tips for business use.
How to charge?
Tips to reduce cost.

Preparation steps
1. Create “Google Cloud Platform(GCP)” account, and
BigQuery.
See) previous paper.
2. Install GCP SKD to your PC. (Using Ubuntu on Vagrant)
1. Installation
2. Activate your account
3. Set accounts for GCP SDK.

2. Install GCP SKD
1. Installation

Install SKD to your PC. (1)
nemo@ubuntu-14:~$ curl https://sdk.cloud.google.com | bash
:
Installation directory (this will create a google-cloud-sdk subdirectory)
(/home/nemo): <-- Just type Enter (or you want)
:
Do you want to help improve the Google Cloud SDK (Y/n)? y
:
! BigQuery Command Line Tool ! 2.0.24 ! < 1 MiB !
! BigQuery Command Line Tool (Platform Specific)! 2.0.24 ! < 1 MiB !
:
Modify profile to update your $PATH and enable shell command
completion? (Y/n)? y (or you want)
:
For more information on how to get started, please visit:
https://cloud.google.com/sdk/#Getting_Started
nemo@ubuntu-14:~$ . ~/.bashrc <-- reload your bash environment
nemo@ubuntu-14:~$

Install SKD to your PC. (2)
// check the commands
nemo@ubuntu-14:~$ which bq
/home/nemo/google-cloud-sdk/bin/bq
nemo@ubuntu-14:~$ which gcloud
/Users/nemo/google-cloud-sdk/bin/gcloud

2. Install GCP SKD
2. Activate your account

Activate your GPC account (1)
1. Preparation (create account)
2. Go to Google Cloud platform (has no account)
3. “Try IT Free”
https://cloud.google.com
nemo@ubuntu-14:~$ gcloud init
Welcome! This command will take you through the configuration of gcloud.
Your current configuration has been set to: [default]
To continue, you must log in. Would you like to log in (Y/n)?
Go to the following link in your browser:
https://accounts.google.com/o/oauth2/auth?redirect_uri=ur&xxxxxxxxxxxxx
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
access_type=offline
Enter verification code: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
You are now logged in as: [xxxx@example.com]
This account has no projects. Please create one in developers console (https://
console.developers.google.com/project) before running this command.
nemo@ubuntu-14:~$

nemo@ubuntu-14:~$ gcloud init
Welcome! This command will take you through the configuration of gcloud.
Your current configuration has been set to: [default]
To continue, you must log in. Would you like to log in (Y/n)?
Go to the following link in your browser:
https://accounts.google.com/o/oauth2/auth?redirect_uri=ur&xxxxxxxxxxxxx
access_type=offline
This account has no projects. Please create one in developers console (https://
console.developers.google.com/project) before running this command.
nemo@ubuntu-14:~$

https://accounts.google.com/o/oauth2/auth?redirect_uri=ur&xxxxxxxxxxxxxxxxx
access_type=offline
Launch
Browser
Select accounts
(if you already login with
multiple accounts)

access_type=offline
Accept
permission

access_type=offline
get
verification code

access_type=offline
set the code

access_type=offline
check the accounts

// set Project ID
nemo@ubuntu-14:~$ gcloud config set project {{PROJECT_ID}}
nemo@ubuntu-14:~$
// check the accounts
nemo@ubuntu-14:~$ gcloud auth list
- xxx@example.com (active)
To set the active account, run:
$ gcloud config set account ``ACCOUNT''
nemo@ubuntu-14:~$

What a pain!
AWS is much easiler...

Try Public data (1)
nemonemo@ubuntu-14:~$ bq show publicdata:samples.shakespeare
Table publicdata:samples.shakespeare
Last modified Schema Total
Rows Total Bytes Expiration
----------------- ------------------------------------
------------ ------------- ------------
26 Aug 21:43:49 |- word: string (required) 164656
6432064
|- word_count: integer (required)
|- corpus: string (required)
|- corpus_date: integer (required)
publicdata : samples . shakespeare
{PROJECT_ID} : {DATASET} . {TABLE}

Try Public data (2)
nemo@ubuntu-14:~$ bq query "SELECT word, COUNT(word) as count FROM
publicdata:samples.shakespeare WHERE word CONTAINS 'raisin' GROUP BY word"
Waiting on bqjob_r5e78fd2c80d5923c_000001554d1c4acc_1 ... (0s) Current
status: DONE
+---------------+-------+
| word | count |
+---------------+-------+
| raising | 5 |
| dispraising | 2 |
| Praising | 4 |
| praising | 7 |
| dispraisingly | 1 |
| raisins | 1 |
+---------------+-------+
nemo@ubuntu-14:~$

Create Dataset (1)
nemo@ubuntu-14:~$ bq ls
<--- no dataset
nemo@ubuntu-14:~$ bq mk saigon_engineers
Dataset 'open-study-group-saigon:saigon_engineers' successfully created.
datasetId
------------------ <-- created!!
saigon_engineers
nemo@ubuntu-14:~$

Create Dataset (2)
<--- no dataset
nemo@ubuntu-14:~$ bq mk saigon_engineers
Dataset 'open-study-group-saigon:saigon_engineers' successfully created.
datasetId
------------------ <-- created!!
saigon_engineers
nemo@ubuntu-14:~$
Added!! -->

Create table and import data (1)
name type
ID INTEGER
name STRING
engineer_type INTEGER
ID name type
1 nemo 1
2 miki 1
Schema
Data

Schema (schema.json)
[
{
"name":"id",
"type":"INTEGER"
},
{
"name":"name",
"type":"STRING"
},
{
"name":"engineer_type",
"type":"INTEGER"
}
]

Data (data.json)
{"id":1,"name":"nemo","engineer_type":1}
{"id":2,"name":"miki","engineer_type":1}

nemo@ubuntu-14:~$ bq load --source_format=NEWLINE_DELIMITED_JSON
saigon_engineers.engineer_list data.json schema.json
Upload complete.
Waiting on bqjob_r23b898932d75d49a_000001554e5cae2f_1 ... (1s)
Current status: DONE
nemo@ubuntu-14:~$
bk load {PROJECT_ID}:{DATASET}.{TABLE} {data} {schema}
Create table and import data
https://cloud.google.com/bigquery/loading-data

saigon_engineers.engineer_list data.json
id:integer,
name:string,
engineer_type:integer
Upload complete.
Waiting on bqjob_r33b7802ea96b2c5d_000001554e4d21d5_1 ... (2s)
nemo@ubuntu-14:~$
Create table and import data : Another way

nemo@ubuntu-14:~$ bq mk open-study-group-
saigon:saigon_engineers.engineer_list schema.json
nemo@ubuntu-14:~$
Create table
bk mk {PROJECT_ID}:{DATASET}.{TABLE} {schema}

saigon_engineers.engineer_list data.json
Upload complete.
Waiting on bqjob_r13717485c2c472e3_000001554e5b3ca3_1 ... (2s)
nemo@ubuntu-14:~$
Import data to database
bk load {PROJECT_ID}:{DATASET}.{TABLE} {data}

Query (1)
nemo@ubuntu-14:~$ bq show saigon_engineers.engineer_list
Last modified Schema Total Rows
Total Bytes Expiration
----------------- --------------------------- ------------
------------- ------------
14 Jun 10:02:35 |- id: integer 2 44
|- name: string
|- engineer_type: integer
nemo@ubuntu-14:~$

Query (3)
nemo@ubuntu-14:~$ bq query --dry_run "SELECT name FROM
saigon_engineers.engineer_list"
Query successfully validated. Assuming the tables are not
modified, running this query will process 12 bytes of data.
nemo@ubuntu-14:~$
bk query --dry_run “QUERY”
- get size of using memory before execution.

Pricing
Storage $0.02 per GB, per month
Long Term Storage $0.01 per GB, per month
Streaming Inserts $0.01 per 200 MB
Queries $5 per TB (First 1 TB per month is free)
subject to query pricing details.
Loading data Free
Copying data Free
Exporting data Free
Metadata operations Free
List, get, patch, update and delete calls.
It seems very cheap !!?

Pricing
Storage $0.02 per GB, per month
Long Term Storage $0.01 per GB, per month
Streaming Inserts $0.01 per 200 MB
Queries $5 per TB (First 1 TB per month is free)
subject to query pricing details.
Loading data Free
Copying data Free
Exporting data Free
Metadata operations Free
List, get, patch, update and delete calls.
BigQuery is for BIG DATA

Column oriented (1)
Sample case : database of Books
ID
(indexed)
title
(indexed)
contents
1 The Cat
Lorem ipsum dolor sit amet,
consectetur (... 1.2MB)
2 Cats are love
3 Littul Kittons
select id, title from books where name = ‘The Cat’

Column oriented (2)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
select * from books where title = ‘The Cat’
@RDBMS
index (name)
hash data
hash data
hash data
data in databaseIndexes
scanned data

Column oriented (3)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
@BigQuery
data in database
scanned data

Column oriented (3)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
@BigQuery
data in database
scanned data
Full-scan 
ANYTIME!!

Column oriented (4)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
@BigQuery
data in database
If your database is Tera-byte scale,
$5 per query !!!!

Column oriented (5)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
select id, title from books where title = ‘The Cat’
@RDBMS
index (name)
hash data
hash data
hash data
scanned data

Column oriented (6)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
@BigQuery
data in database
scanned data

Column oriented (6)
ID
(indexed)
title
(indexed)
contents
1 The Cat
2 Cats are love
3 Littul Kittons
@BigQuery
data in database
scanned data
Column
Oriented

It's really
dangerous!
Please, Please set columns in queries.

Table division
Sample case : database of Books
select id, title from books where time in ‘2016/06/17’
: : : :
ID
(indexed)
title
(indexed)
contents
time
(indexed)
1 The Cat
2016/01/01
00:00:00
2 Cats are love
2016/01/01
00:01:23
353485397 Littul Kittons
2016/06/17
00:01:46

Table division (1)
index (time)
hash data
hash data
hash data
scanned data
: : : :
ID
(indexed)
title
(indexed)
contents
time
(indexed)
1 The Cat
Lorem ipsum dolor sit
amet, consectetur (...
2016/01/01
00:00:00
2
Cats are
love
2016/01/01
00:01:23
353485397
Littul
Kittons
0.8MB)
2016/06/17
00:01:46
@RDBMS

Table division (2)
data in database
scanned data
: : : :
ID
(indexed)
title
(indexed)
contents
time
(indexed)
1 The Cat
1.2MB)
2016/01/01
00:00:00
2 Cats are love
1.5MB)
2016/01/01
00:01:23
353485397
Littul
Kittons
0.8MB)
2016/06/17
00:01:46
@BigQuery
Huge size

Table division (3)
ID
(indexed)
title
(indexed)
contents
time
(indexed)
1 The Cat
1.2MB)
2016/01/01
00:00:00
2 Cats are love
1.5MB)
2016/01/01
00:01:23
ID
(indexed)
title
(indexed)
contents
time
(indexed)
353485397
Littul
Kittons
0.8MB)
2016/06/17
00:01:46
:
Tables
books_20160101
:
books_20160617
Divide tables for each day.

Table division (4)
ID
(indexed)
title
(indexed)
contents
time
(indexed)
1 The Cat
1.2MB)
2016/01/01
00:00:00
2 Cats are love
1.5MB)
2016/01/01
00:01:23
ID
(indexed)
title
(indexed)
contents
time
(indexed)
353485397
Littul
Kittons
0.8MB)
2016/06/17
00:01:46
:
books_20160101
:
books_20160617
@BigQuery

Table division (5)
ID
(indexed)
title
(indexed)
contents
time
(indexed)
1 The Cat
1.2MB)
2016/01/01
00:00:00
books_20160101
::
ID
(indexed)
title
(indexed)
contents
time
(indexed)
353485397
The Great
Catsby
0.8MB)
2016/06/16
00:01:46
books_20160616
select id, title from books
where time in ‘2016/06/16 - 2016/06/17’
@BigQuery
ID
(indexed)
title
(indexed)
contents
time
(indexed)
353485397
Littul
Kittons
2016/06/17
00:01:46
books_20160617

Table division (6)
select id, title from books
where time in ‘2016/06/16 - 2016/06/17’
@BigQuery
SELECT id, title FROM
(
TABLE_DATE_RANGE(books_,
TIMESTAMP(‘2016-06-16'),
TIMESTAMP(‘2016-06-17')
)
)

Table division (7)
Other ways to divide tables.
Table decorator  
- https://cloud.google.com/bigquery/table-decorators
“TABLE_QUERY”
- https://cloud.google.com/bigquery/query-reference
“Import from GCS is much faster than from local”
1. put data into GCS (Google Clould Storage ≒ S3 ??)
2. import the data from GCS.
Other tips.

BigQuery is
Fast
Easy
Cheap
if it is used properly.

BigQuery is
Fast
Easy
Cheap
if it is used properly.
Remember
“--dry_run”

Big query - Command line tools and Tips - (MOSG)

More Related Content

What's hot (20)

Similar to Big query - Command line tools and Tips - (MOSG) (20)

More from Soshi Nemoto (10)

Recently uploaded (20)

Big query - Command line tools and Tips - (MOSG)