Contents：
In this section, you will create two functions to implement simple WordCount, and test the functions through the console or by calling API.
Creating a Mapper Function
Creating a function through the console
1) Log in to the Serverless Cloud Function Console. Select Guangzhou from the region list and click Create.
2) In the Function configuration section, enter Mapper as the function name and keep default settings for the other options, and then click Next.
3) Go to the Function code section, and click Local upload zip file. Enter map_function.main_handler for the execution method, select the mapper.zip created in "Step 2: Create Deployment Package", and click Next.
4) In the Triggering method section, you need to test the function manually, so no trigger method is added. Click Done.
Creating a function using API
For more information, please see API CreateFunction.
Creating a Reducer Function
Creating a function through the console
1) Log in to the Serverless Cloud Function Console. Select Guangzhou from the region list and click Create.
2) In the Function configuration section, enter Reducer as the function name, leave all the other configuration options unchanged, and then click Next.
3) Go to the Function code section, and click Local upload zip file. Enter reduce_function.main_handler for the execution method, select the reducer.zip created in "Step 2: Create Deployment Package", and click Next.
4) In the trigger method section, you need to test the function manually, so no trigger method is added. Click Done.
Creating a function using API
For more information, please see API CreateFunction.
Testing Functions
When a function is created, it is generally tested through the console or API, to ensure the function output meets the expectation, and then you can bind it to a trigger for practical application.
Testing the function through the console
1) In the details page of the Mapper function you just created, click Test.
2) Choose Upload File to COS/Delete File from COS Test Code from the drop-down list of test templates.
3) In the test code, set name to the name of bucket srcmr created in "Step 1: Prepare COS Bucket", and set key to the key value of /serverless.txt uploaded in "Step 1: Prepare COS Bucket", as shown in the example below:
{
"Records":[
{
"event": {
"eventVersion":"1.0",
"eventSource":"qcs::cos",
"eventName":"event-type",
"eventTime":"Unix timestamp",
"eventQueue":"qcs:0:cos:gz:1251111111:cos",
"requestParameters":{
"requestSourceIP": "111.111.111.111",
"requestHeaders":{
"Authorization": "Uploaded authentication information"
}
}
},
"cos":{
"cosSchemaVersion":"1.0",
"cosNotificationId":"Configured or returned ID",
"cosBucket":{
"name":"srcmr", #set to demo bucket here
"appid":"appId",
"region":"gz"
},
"cosObject":{
"key":"/serverless.txt", #set to demo file here
"size":"1024",
"meta":{
"Content-Type": "text/plain",
"x-cos-meta-test": "Custom meta",
"x-image-test": "Custom meta"
},
"url": "Origin server URL for accessing files"
}
}
}
]
}4) Click Run to view the results.
5) Go to the COS Console, and click destmr created in "Step 1: Prepare COS Bucket", to check whether a file named result_middle_serverless.txt is generated in COS Bucket. The file collects the number of times each word appears in the uploaded text file:
6) Download this file, and you can see the content similar to the following:
the 29
as 25
to 20
of 17
and 16
a 16
code 16
in 15
be 14
or 12
serverless 10
is 10
that 10
computing 9
by 9
for 9
an 8
not 8
cloud 7
functions 6
edit 6
can 6
with 5
servers 5
on 5
because 5
at 5
Serverless 5
have 5
platform 5
such 5
time 4
virtual 4
up 4
function 4
provider 4
autoscaling 4
used 4
it 4
js 4
are 4
also 4
than 4
service 4
written 4
2016 4
1 4
runtime 4
example 4
use 4
Node 4
run 3
does 3
Java 3
execution 3
other 3
model 3
In 3
requests 3
means 3
typically 3
which 3
latency 3
any 3
could 3
This 3
may 3
via 3
resources 3
required 2
10 2
developers 2
application 2
underutilisation 2
hosted 2
was 2
start 2
support 2
microservices 2
per 2
back 2
server 2
source 2
generally 2
It 2
well 2
user 2
running 2
Python 2
handling 2
JSON 2
addition 2
limits 2
both 2
available 2
For 2
request 2
The 2
efficient 2
part 2
2 2
system 2
APIs 2
first 2
completely 2
supports 2
container 2
programmer 2
API 2
name 2
provision 2
now 2
they 2
3 2
its 2
operations 2
machine 2
end 2
more 2
from 2
public 2
officially 2
even 2
cost 2
Functions 2
all 2
requires 1
At 1
2006 1
starting 1
serialized 1
calls 1
assets 1
InterConnect 1
exposure 1
needed 1
included 1
suited 1
includes 1
you 1
5 1
setting 1
when 1
no 1
periods 1
abstract 1
defined 1
called 1
continuously 1
simplifying 1
Infrequently 1
but 1
triggered 1
significantly 1
registration 1
down 1
functional 1
into 1
own 1
software 1
automatically 1
introduced 1
about 1
although 1
processing 1
business 1
However 1
latencies 1
some 1
creation 1
exposed 1
spins 1
either 1
specific 1
necessary 1
using 1
rules 1
usable 1
has 1
Despite 1
4 1
web 1
world 1
being 1
just 1
spend 1
without 1
person 1
task 1
provisioned 1
uses 1
Resource 1
meets 1
development 1
composition 1
units 1
make 1
pay 1
if 1
worry 1
technology 1
launched 1
stopping 1
owns 1
policies 1
will 1
batch 1
unlike 1
case 1
demand 1
ensuring 1
bulk 1
containers 1
offers 1
been 1
billing 1
tuning 1
GitHub 1
sequences 1
simply 1
simple 1
point 1
consume 1
Monitoring 1
s 1
sensitive 1
triggers 1
essentially 1
packages 1
premise 1
Programming 1
initially 1
terms 1
C 1
invoker 1
behind 1
capacity 1
Performance 1
configured 1
wrote 1
where 1
fixed 1
architecture 1
imposed 1
Swift 1
resource 1
infrequent 1
conjunction 1
cheaper 1
do 1
8 1
black 1
only 1
outside 1
renting 1
quantity 1
Cloud 1
Cost 1
traditional 1
considered 1
alpha 1
significant 1
performance 1
online 1
purchasing 1
responsible 1
offering 1
option 1
need 1
rent 1
debugging 1
jobs 1
PaaS 1
11 1
specialists 1
released 1
serve 1
Zimki 1
structures 1
their 1
management 1
given 1
open 1
directly 1
instances 1
number 1
features 1
major 1
launch 1
language 1
multithreading 1
announced 1
7 1
since 1
measure 1
another 1
events 1
fully 1
9 1
introduce 1
Alternatively 1
greater 1
affected 1
non 1
might 1
2014 1
systems 1
production 1
actually 1
group 1
style 1
suffer 1
including 1
order 1
allow 1
Haskell 1
purchase 1
response 1
data 1
designed 1
produce 1
REST 1
provisioning 1
rather 1
involve 1
providers 1
dedicated 1
believed 1
this 1
hour 1
endpoint 1
high 1
billed 1
known 1
expose 1
sharing 1
6 1
Microsoft 1
would 1
environment 1
private 1
followed 1
FaaS 1
involves 1
amount 1
deserialized 1
box 1
groups 1
supporting 1
satisfy 1
new 1
version 1
Disadvantages 1
machines 1
likely 1
cron 1
HTTP 1
workloads 1
small 1
level 1
announcing 1
include 1
languages 1
Docker 1
go 1
manages 1