C++ Driver showing "Failed to create prepared statement on table sdp01.Request timed out"

[Deepak Kumar]

Hi All,

I am facing an issue while running my application.
I am using a 3 Node Cassandra cluster. And created a keyspace with RF 2 and table sdp01 inside it.

My application is running 16 processes with 4 thread each. Threads inside one process shares CassSession.

In my application for each cassandra statement, i am setting CONSISTENCY LEVEL as ONE.

I started my application and everything is running fine. Data is retrieved from cassandra.

But the problem appears when, I kill cassandra server on one of the nodes.
Now for some select queries, above mentioned error is visible. For some select queries this is also visible "Failed to execute database query because "Request timed out""

Why this error is coming, since the consistency level is ONE, shouldn't all select queries should pass as 2 Nodes are still up and RF is 2?

I am stuck in this problem and not able to find any reason for this behaviour, one probable reason(theoratically) can be, that, IO thread has made connection with the node which I has shut down, and driver is yet not aware that this node is down. This might result in timeout for those select queries, untill cassandra driver is aware of the node down.

Please help here......

[Michael Penick]

If your application's queries (as are SELECTs queries) are idempotent it can use `cass_statement_set_is_idempotent(statement, cass_true)` to tell the driver that it's safe to retry those queries when connections fail.

More information can be found here: https://github.com/datastax/cpp-driver/blob/master/include/cassandra.h#L4048-L4062

Mike

--
You received this message because you are subscribed to the Google Groups "DataStax C++ Driver for Apache Cassandra User Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cpp-driver-user+unsubscribe@lists.datastax.com.