Easy Deployment and Operations

About This Topic

Systems don't end when they're built—continuing to operate them is what matters. This project focused on designs that minimize effort from deployment through operations.

Auto-Deployment to Cloud

Just upload source code, and it's automatically deployed to the cloud environment.

Deployment Flow

Push Code

Developer sends code to Git repository

Auto-Detection

Cloud platform detects changes

Build

Install dependencies, compile

Deploy

Roll out new version

Health Check

Verify normal operation and complete

Auto-Deployment Benefits

Benefit	Description
No server provisioning	No physical servers or VPS needed
Reduced setup effort	No OS or middleware configuration
Immediate reflection	Minutes from code push to production
Rollback capability	Can revert to previous version if issues

No server provisioning

DescriptionNo physical servers or VPS needed

Reduced setup effort

DescriptionNo OS or middleware configuration

Immediate reflection

DescriptionMinutes from code push to production

Rollback capability

DescriptionCan revert to previous version if issues

Auto-Scaling

Even as usage increases, the cloud automatically adjusts resources.

Scaling Mechanism

Quiet Period

Minimal configuration for cost savings. 1 instance handles processing

Access increases

Normal Period

2-3 instances process in parallel

Month-end busy period

Busy Period

Automatically scale out. Distribute request processing

Scaling Points

Situation	Instance Count	Response
Night/Weekend	Minimum (0-1)	Minimize cost
Normal business hours	2-3	Stable response
Month-end aggregation	Auto-increase	Prevent processing delays

Night/Weekend

Instance CountMinimum (0-1)

ResponseMinimize cost

Normal business hours

Instance Count2-3

ResponseStable response

Month-end aggregation

Instance CountAuto-increase

ResponsePrevent processing delays

Serverless architecture means you pay only for what you use. No charges during periods without access, making it cost-effective.

Configuration via Environment Variables

Settings are managed through environment variables. Adjust behavior without changing code.

Required Settings

Variable	Description	Example
RUN_TOKEN	Secret token for authentication	32+ character string
FREEE_CLIENT_ID	Invoicing service client ID	ID issued by API
FREEE_CLIENT_SECRET	Invoicing service client secret	Secret key issued by API
FREEE_REFRESH_TOKEN	Refresh token for token renewal	Token issued by API

RUN_TOKEN

DescriptionSecret token for authentication

Example32+ character string

FREEE_CLIENT_ID

DescriptionInvoicing service client ID

ExampleID issued by API

FREEE_CLIENT_SECRET

DescriptionInvoicing service client secret

ExampleSecret key issued by API

FREEE_REFRESH_TOKEN

DescriptionRefresh token for token renewal

ExampleToken issued by API

Optional Settings (with defaults)

Variable	Description	Default
RATE_LIMIT_PER_MINUTE	Request limit per minute	60
CONCURRENT_FREEE_CALLS	Concurrent API calls	2
RETRY_MAX_ATTEMPTS	Maximum retry attempts	5
REQUEST_TIMEOUT_MS	Timeout (milliseconds)	30000

RATE_LIMIT_PER_MINUTE

DescriptionRequest limit per minute

Default60

CONCURRENT_FREEE_CALLS

DescriptionConcurrent API calls

Default2

RETRY_MAX_ATTEMPTS

DescriptionMaximum retry attempts

Default5

REQUEST_TIMEOUT_MS

DescriptionTimeout (milliseconds)

Default30000

Testing Settings

Variable	Description	Purpose
FREEE_DRY_RUN	Dry-run mode (true/false)	Test without production send
ENABLE_STRUCTURED_LOGGING	Detailed log output (true/false)	Enable during debugging

FREEE_DRY_RUN

DescriptionDry-run mode (true/false)

PurposeTest without production send

ENABLE_STRUCTURED_LOGGING

DescriptionDetailed log output (true/false)

PurposeEnable during debugging

Timeout Settings

Appropriate timeouts are set for long-running processes.

Timeout Control

Request Starts

Begin processing, set timer

Execute Processing

Run API calls, etc.

Timeout Check

Exceeded set time?

Within time

Return result normally

Exceeded

Abort processing, return error

Timeout Setting Guidelines

Processing	Recommended	Reason
Single API call	30 seconds	Usually completes in seconds
Batch (~100 items)	2 minutes	Account for paging
Large scale (~500 items)	5 minutes	Use cloud's full limit

Single API call

Recommended30 seconds

ReasonUsually completes in seconds

Batch (~100 items)

Recommended2 minutes

ReasonAccount for paging

Large scale (~500 items)

Recommended5 minutes

ReasonUse cloud's full limit

Distributed Cache Utilization

Using Redis (Upstash) distributed cache enables:

Function	Description	Benefit
Idempotency cache	Keep processed keys for 7 days	Shared across instances
Token cache	Keep access token until expiry	Reduce API calls
Distributed lock	Prevent concurrent updates	Ensure data consistency

Idempotency cache

DescriptionKeep processed keys for 7 days

BenefitShared across instances

Token cache

DescriptionKeep access token until expiry

BenefitReduce API calls

Distributed lock

DescriptionPrevent concurrent updates

BenefitEnsure data consistency

Distributed cache is optional but recommended. Without it, only memory cache operates, but cache isn't shared between instances, risking duplicate processing.