Elasticsearch is your way to go!
Use Cases and Deployment Scope
Elasticsearch is an important service that we use frequently in the organization. We use Elasticsearch as a logging service for our system logs, Once we have the logs in Elasticsearch, we connect to Kibana and start building dashboards and charts that help us track our system stability and availability in terms of System metrics. On the other hand, we use it to track new bugs and errors. The other usage for Elasticsearch in our system is as a search engine. Elasticsearch is a very fast and amazing search engine, where we store some fields and call Elasticsearch APIs to fetch these fields when needed.
Pros
- Log management
- Search Engine
- Autocomplete service
- Storing Data
- Caching layer in some cases
- ML and data analysis
Cons
- Elasticsearch is kind of hard to maintain as a cluster on k8s when self-managed.
- Good to support AI that will help buidling complex queries
- Documentation for Java library of Elasticsearch and Elasticsearch client is not that great compared to the APIs documentation
Return on Investment
- We're able to detect system incidents at early stage
- We can achieve 99.98% availability for searching service
- Data retrieval speed is less than 300 ms on avg
Alternatives Considered
Sumo Logic, Splunk Log Observer and Prometheus
Other Software Used
Splunk On-Call, ManageEngine Site24x7, MongoDB, Redis™*, Snyk, Cloudflare, Domo, Slack, Fullstory, NetSuite ERP, LaunchDarkly, Ghost Inspector
