I am new to cassandra and migrating my application from Mysql to cassandra.As i read of cassandra it says it reduces the read and write time compared to Mysql.when i tried a simple example with single node using hector, reading operation is quite faster compared to Mysql,and if i tried to insert a single column it is very fast compared to mysql.But when i tried to insert a single row with multiple columns its taking long time compared to mysql. Is there anyway to improve Write performance or please let me know if i am wrong with my way of coding.
My sql code is
INSERT into energy_usage(meter_id,reading_datetime,reading_date,reading_time,asset_id,assetid)
VALUES('164','2012-12-07 00:30:00','2012-12-07','00:00:00','1','1') "
my cassandra code is
Mutator<String> mutator = HFactory.createMutator(keyspaceOperator, StringSerializer.get());
mutator.addInsertion("888999", DYN_CF, HFactory.createStringColumn("assetid","1")).
addInsertion("888999", DYN_CF, HFactory.createStringColumn("meterid", "164")).
addInsertion("888999", DYN_CF, HFactory.createStringColumn("energyusage","10")).
addInsertion("888999", DYN_CF, HFactory.createStringColumn("readdate","2012-12-07")).
addInsertion("888999", DYN_CF, HFactory.createStringColumn("readdatetime","2012-12-07 00:30:00"));
mutator.execute();
Well, if you can provide us with some benchmarks stating what do you find as 'very slow', maybe we can help you further. But one important thing is to note that a cassandra feature is that as the amount of data in your cluster grows, you are guaranteed (loosely) to get somewhat more constant read/write performance times as compared to a relational db engine.
Have you tried executing the same query in a cassandra-cli/cqlsh, for example? What kind of performance do you get then? If the difference between the cassandra-cli/cqlsh statement and the piece of code you posted is dramatic, that means that it is probably a driver issue, so I would dig deeper in Hector code.