Solution Architecture - Evaluations & Selections

Thank you for reading my latest article here.

Here at LinkedIn, I regularly write about data architecture, Business Architectures Business Concepts and technology trends. To read my future articles simply join my Newsletter on LinkedIn or follow me on 'Rohan Rekhi | LinkedIn

Background

I have been covering various Pillars of Solution Architecture in last few articles. For each of these pillars we do need solutioning and there are usually multiple options to choose from. A solution selection and evaluation is series of recommendations and/or restrictions for all potential viable combinations of architectural standards, technologies or platforms to be used in the design and implementation of a technology solution. This artifact type is determined by declaring predefined dispositions, it provides clarity on the preferred selections within a solution.?

We need Solution Selection & Evaluation in the direct and viable choices in particular technology categories. Each solution selector contains the different technologies available for the category in question, their current lifecycle state (current, legacy, planned, etc) and is arranged in a way that enables one to choose the appropriate implementation based on their use case. Prerequisite to using one of these technologies is that the technology must be present and described in the corresponding architectural blueprint. If the technology is not described in said blueprint, please engaged the Enterprise Architects for the area in question to determine a path forward.

Few Technology categories listing here, list may not be exhaustive.

1. API Gateway solution
2. Authentication solution
3. Authorization solution
4. Batch solution
5. Caching solution
6. Cross-environment file share solution
7. On-prem Database Platform solution
8. Cloud Database Platform solution
9. On-prem Database Performance Benchmark
10. Cloud Database Performance Benchmark
11. Performance Testing solution
12. Functional Test Automation Framework solution
13. Cloud Compute solution
14. Hosting platform solution
15. Managed File Transfer solution
16. Storage solution
17. Reference Architecture solutions
18. Technology Targets

Lets look at Cloud Database Platform Solution briefly as that is one of the big decision point of any Solution Architecture.

For Public Cloud Provider usage guidelines should be documented..?All applications use cases that needs to use Public Cloud needs to follow the Governance which includes review and approval from Enterprise Architecture Authorities. Not all the cloud database services may be ready for supporting productions workloads and evaluation and design pattern development might still be ongoing for many of them.?The Cloud Database solution selection document should be frequently updated (6 to 9 Months), as the evaluation and design pattern development of the Database services offered by Public Cloud providers are completed.

Lets look at potential Current and Future state indicative considerations for Google GCP and Amazon AWS platforms.

GCP

Current State

1. Cloud SQL PostgreSQL is the?recommended?default database platform for new?and existing applications that requires relational data store for structured data durability in Google Cloud.
2. Cloud SQL PostgreSQL is also recommended for applications that requires schema flexibility (via use of JSONB features).
3. Use cases that requires read scalability and are tolerant of eventually consistent reads should use Cloud PostgreSQL Read Replicas to achieve read scaling.
4. Cloud SQL SQL Server can be the specialty database platform for existing applications that uses SQL Server on-prem and? has to migrate to Google cloud but can't re-write application to work with preferred platform(s)?
5. MongoDB Atlas is the specialty Document DB platform for existing applications that uses MongoDB on-prem and has to migrate to Google Cloud Platform but can't re-write application to work with preferred platform(s).

Future State

1. Cloud Spanner will be the specialty database platform for the new and existing?applications? that requires relational data store for structured data durability and also requires horizontal scalability or strongly consistent transactions across regions
2. Cloud Firestore will be the specialty?Document DB for?new and existing? applications that requires schema flexibility and horizontal write/strongly consistent read scalability (Enhanced Point In Time recovery on roadmap for 2023)
3. Aerospike Managed service (roadmap offering) will be the specialty DB platform for existing applications that uses Aerospike on-prem?and has to migrate to Google Cloud Platform but can't re-write application to work with preferred platform(s)

AWS

Current State

1. AWS Aurora PostgreSQL is the recommended?database platforms for new? and existing applications that requires relational data store for structured data durability in AWS Cloud Platform.
2. AWS Aurora?PostgreSQL is also recommended for applications that requires schema flexibility (via use of JSON features).?
3. Use cases that requires read scalability and are tolerant of eventually consistent reads should use Aurora PostgreSQL Read Replicas to achieve read scaling.?
4. AWS DynamoDB is the recommended Document DB for?new? and existing?applications that requires schema flexibility and horizontal scalability
5. AWS Aurora MySQL is the specialty?database platforms for existing?applications that requires relational data store for structured data durability in AWS Cloud Platform

Future State

1. AWS RDS SQL Server will be the specialty database platform for?existing applications that uses SQL Server on-prem and has to migrate to AWS cloud but can't re-write application to work with preferred?platform(s)
2. MongoDB Atlas will be the specialty Document DB platform for existing applications that uses MongoDB on-prem?and has to migrate to AWS Cloud Platform?but can't re-write application?to work with preferred platform(s)

ANY request for a new data store MUST be accompanied by capacity, growth, and performance models, along with RTO and RPO targets to ensure the correct cluster configuration is chosen.

Public Cloud Persisted Database patterns

Online Transaction Processing (AKA OLTP) databases are used to?provide persisted, durable data storage for applications needing to respond immediately to user requests.? This includes any application databases classified as systems of record.

Here's a quick snippet of the AWS and GCP typical offerings

In future articles, I will zoom into few other Selection Technology categories.

#solutionarchitecture #selection #evaluation #performance #resilliancy #scalability #serverless

To stay up to date with my latest articles in, make sure to subscribe to my newsletter follow me on LinkedIn, and if you or anyone in your network is interested in taking a deeper dive into some of these topics or looking for help with your initiatives and programs, please feel free to reach out to me. For wider reach please share.

You can also follow me on Medium and Subscribe to my articles there.