Default sizes for microservices containers

Adeptia Connect allows you to configure the CPU and memory for a microservice by setting the following parameters in the values.yaml file.

• CPU and Memory limit - The maximum number of CPU cores and the memory a microservice pod can consume.
• CPU and Memory Request - The minimum number of CPU cores and the memory required by a microservice pod.
• Minimum Replicas - The minimum number of replicas (pods) of a microservice that will run.
• Maximum Replicas - The maximum number of replicas (pods) to which a microservice can scale up.

The table below details the technology stack and resource requirements for each microservice in the deployment. It helps you plan the CPU and memory strategies based on individual service needs.

Key information includes:

• Technology Stack used for each service (e.g., Java Spring Boot, RabbitMQ, Python).

• Java options (Java opts) like heap size (Xms, Xmx) and runtime flags for JVM-based services.

• CPU and Memory specifications:

  • CPU Request and CPU Limit

  • Memory Request and Memory Limit

• Replicas defined for:

  • Default setup

  • High Availability (HA) setup

Service	Technology Stack	Java opts	CPU Request	CPU Limit		Memory Request	Memory Limit	Replicas (default)	Replicas (HA)
runtime	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms1024M -Xmx4096M -Djdk.tls.useExtended MasterSecret=false	0.4 CPU	2 CPU		1638Mi	8192Mi	Min: 2	Min: 2
								Max: 2	Max: 2
rabbitmq	RabbitMQ v3.13.7	Not applicable	0.2 CPU	1 CPU		820Mi	4096Mi	Min: 1	Min: 1
								Max: 1	Max: 3
event	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms1024M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.2 CPU	1 CPU		820Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2
listener	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms1024M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	1 CPU		410Mi	4096Mi	Min: 1	Min: 1
								Max: 1	Max: 1
api-gateway	Java Spring Cloud Gateway	Xms512M -Xmx1024M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	1 CPU		410Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2
soap-api	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	0.8		410Mi	3276Mi	Min: 1	Min: 2
								Max: 1	Max: 2
rest-api	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	1 CPU		410Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2
webapp-gateway	Java Spring Cloud Gateway	Xms512M -Xmx1024M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	0.8		410Mi	3276Mi	Min: 1	Min: 2
								Max: 1	Max: 2
portal	Ext-JS v7.3.0 (JavaScript library) BPMN-JS v0.27.6 (JavaScript library for Web PD)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.2 CPU	1 CPU		820Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2
web-runner	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	1 CPU		410Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2
license	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	0.8		410Mi	3276Mi	Min: 1	Min: 2
								Max: 1	Max: 2
archival-and-cleanup	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	0.8		410Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2
deploymentmanager	Shell Script	Not applicable	0.3 CPU	1.5 CPU		410Mi	3276Mi	Min: 1	Min: 2
								Max: 1	Max: 2
aimap	Python v3.6	Not applicable	0.1 CPU	1 CPU		410Mi	3276Mi	Min: 1	Min: 1
								Max: 1	Max: 2
redis (optional)	Redis v7.4.1	Not applicable	0.4 CPU	0.8		400Mi	1Gi	Min: 3	Min: 3
								Max: 3	Max: 3
sapservice (optional)	Java Spring Boot v3.5.12 (Embedded Tomcat)	Xms512M -Xmx2048M -Djdk.tls.useExtended MasterSecret=false	0.1 CPU	1 CPU		410Mi	4096Mi	Min: 1	Min: 2
								Max: 1	Max: 2

Total resource requirements

This section summarizes the total resource requirements for deploying the application in both Default and High Availability (HA) configurations.

Key highlights:

• Covers both Default and HA setup scenarios.

• Includes resource requirements for:

  • Core microservices

  • Core microservices with Redis service enabled

  • Core microservices with SAP service enabled 

  • Core microservices with both Redis and SAP service enabled 

• Includes requirement details for:

  • CPU Cores: Total number of recommended CPU cores

  • VMs: Total recommended number of VMs, calculated based on the total recommended CPU cores 

  • Memory Request: Total minimum memory required

  • Memory Limit: Total maximum memory that can be consumed​

  • CPU Request: ​Total minimum number of CPU cores required 

Total resources requirements	Default /HA setup	Core microservices	With Redis enabled	With SAP service enabled	With both Redis and SAP service enabled
CPU Cores (Calculated as: Sum of [CPU Limit x Min Replicas])	Default	16.7	19.1	17.7	20.1
	HA	26.4	28.8	28.4	30.8
VMs	Default	3 VMs each with 8 CPU cores (Using 8-core CPU Recommended)
	HA	4 VMs each with 8 CPU cores (Using 8-core CPU Recommended)
Memory Request	Same for default and HA setup	8198Mi	8598Mi	8608Mi	9008Mi
Memory Limit	Same for default and HA setup	57340Mi	58364Mi	61436Mi	62460Mi
CPU Request	Same for default and HA setup	15.5	15.9	15.6	16

<ch:anchor>It is recommended that you use the 8-core VMs for the following reasons:</ch:anchor> An 8-core VM allows Kubernetes to dynamically allocate resources more efficiently, reducing fragmentation and underutilization. A larger memory pool in an 8-core VM prevents wastage and uneven allocation, reducing the risk of pod evictions due to resource constraints. It also minimizes cross-VM latency, benefiting microservices that require high memory usage.

You can set the maximum number of replicas for a microservice only after enabling its autoscaling property by setting the value for enabled to true in the values.yaml file. By default, autoscaling is disabled. After enabling the autoscaling, you can set the value for maxReplicas variable based on your requirement. Given below is an example showing autoscaling settings for the webrunner microservice. 

The maximum number of replicas for the listener and license microservices cannot be more than what has been specified in the respective table discussed above.

Though Adeptia recommends the maximum number of replicas for each microservice, it may vary (except for the listener and license microservices) based on your requirement.

Microservices at a glance

Component	Description
webapp-gateway	Microservice to serve as a single entry point for communicating with all the Microservices Invokes multiple Microservices and aggregates the results.
portal	Microservice for serving GUI to configure entities and view logs.
web-runner	Microservice for serving UI data.
event	Microservice for scheduling of events
runtime	Microservice for Process Flow and Transaction execution
rest-api	Microservice to publish and serve all the REST API
soap-api	Microservice to publish and serve all the SOAP API
listener	Microservice for the execution of listeners (Only one instance/Pod will run at a time). The following listeners are part of this microservice. For more details on how to work with these listeners, refer to their respective pages. JMS Spazio MFT MLLP SAP Kafka
ai-map	Microservice for AI-based auto mapping.
license	Microservice to register and enforce licensing.
api-gateway	Microservice to serve as a single entry point for all the REST and SOAP API calls.
archival-and-cleanup	Microservice for log cleanup and archival.
rabbitmq	For facilitating asynchronous inter-service communication. Back pressure handling: queueing jobs, routing to dedicated pods, Autoscaling based on queue size.
sapservice (Optional)	Microservice for enabling integration with SAP systems by handling data exchange, communication protocols, and processing logic required for SAP-specific operations. This is by default disabled and can be enabled while you deploy the application.
redis (Optional)	Microservice for enabling distributed execution of Child Process Flows.  This is by default disabled and can be enabled while you deploy the application.