HomeValue Search
- Current Flow
- Problems
- Ideal Flow
- Options
- Replication Strategies
- Implementation Strategies
Current Flow
Current Flow
Problems
Problems
Database address matching is done by GooglePlacesId
Search is bounded by API threshold
Google address database is treated as the source of truth database
Ideal Flow
Ideal Flow
HomeValue DB
HomeValue Identifiers
Options
Options
Managed
Not Managed
AutoComplete frontend libs
Friendly graphic user interface
Managed (Very little manage overhead)
IP based location searches
Intuitive API and API documentation
Isolated Search Database
Metrics and Stats Dashboard
Ingestion
HTTP API
Same ecosystem as application (AWS)
Rest API
Managed by AWS
Intuitive GeoSpatial querying
Ingestion
AWS Firehose
HTTP API
Highly configurable
Rest API
Plugin Architecture
GeoSpatial Querying
Metrics and Stats Dashboard
Ingestion
HTTP API
Elastic Search needs an ElasticSearch-jdbc connector installed. The jdbc connector performs the data ingestion.
Hosted
HTTP API
AWS Firehose
Integrates well with other Apache software right out of the box
Highly configurable
Ingestion
Built in support for ingestion
Ideal Flow
No Google!
HomeValue Identifier
data replication strategies ?
implementation strategies ?
Can the product function ?
Is search an important component of HomeValue ?
How much time we want to allocate towards search functionality ?
Do we want to cross the bridge when we get there ?
How much search volume should we expect ?
What features do we need ?
- Facets
- Word Weight Ranking
- Geospatial querying
What ecosystem should the product belong in ?
- Apache
- AWS
- Other
HomeValue Search
By Milos Bejda
