CODEWITHSUNDEEP
node.jsdatabasecassandra
unnamed-bull
unnamed-bull

Reputation: 361

Best approach to store historical data?

I am using cassandra as database with a table "jobhistory" partitioned based on timestamp . I have 2 million rows in jobhistory table and it can be increasing over time.

My Implementation: I am purging rows in jobhistory table every 7 days ie setting row level TTL value of 7 Days

Can any one suggest me whether to store in jobhistory table and purge it every 7 days or save it log file using node.js and winston logger and do log rotation , which is the best way to solve this scenario. Please, kindly give the supporting reason as to why it is better .

Any better solutions are also welcome. Any help is appreciated

Upvotes: 1

Views: 382

Answers (1)

sandeep rawat
sandeep rawat

Reputation: 4957

It will be good if you put TTL while inserting data in cassandra

ie

INSERT INTO kespace.tablename (c1, c2, c3, time) VALUES (2,'Hello','Mary',toTimestamp(now())) USING TTL 30;

Upvotes: 3

Related Questions