Spike: Types compatibility and Codecs about troy HOT 3 CLOSED

1. Alter Table's alter type only allows very limited type changes, for example ( int -> varint )
2. row.getVarint(...) on a column with a type int fails
3. row.getVarint(...) on a column that has been changed from int to varint, but had existing data, still fails! (Why Cassandra allowed it first place?)

IOError: java.io.EOFException: EOF after 64 bytes out of 2148938))
com.datastax.driver.core.exceptions.NoHostAvailableException: All host(s) tried for query failed (tried: /127.0.0.1:9042 (com.datastax.driver.core.exceptions.ServerError: An unexpected error occurred server side on /127.0.0.1:9042: java.io.IOError: java.io.EOFException: EOF after 64 bytes out of 2148938))

from troy.

Some types are exactly the same, like text and varchar, it doesn't matter anything changing between them.

from troy.

Conclusion: Altering column types would definitely break Parsing if compiled on an older schema (and vice versa), even if types were compatible like (int and varint), but would still work for types were aliases like text and varchar.

Allowing parsing across different schemas requires more work, to be researched later.

from troy.