How to structure and map my data in ElasticSearch?

[moe.kazemi (Moe)]

I am very new with Elasticsearch & I want to use ELK to bring one our reports from SSMS and use Elasticsearch to expedite the processing time.

In our current design in SSMS we have 3 store proc (Master, Report & Result) which fetch data from 10 tables inside our data ware house.

Questions :

1. what is the best way to use ELK to have the same result that we have right now?

2. Is it better to create one store porc and query that with logstash to fetch data from all those tables?

3. or just query each table separately with simple select statement in logstash and fetch data to elsticsearch ?

4. if I do fetch each table separately how I would be able to connect each table on ElasticSearch?

5. After having all the data in ElasticSearch How to Structure the relationship of our data? What is the Best Way to structure?

6. Because this report is same for 4 different clients, Should I Have separate index for each client? Or have them all in one index as for that specific report?

Thank you

[warkolm (Mark Walkom)]

What's SSMS?

[moe.kazemi (Moe)]

My Bad it is, SQL Server Management Studio

[warkolm (Mark Walkom)]

Righto. Then you can probably use Logstash + it's JDBC input to grab data from it. That way you can build your report with a SQL query and that will get translated into a json document into Elasticsearch.