Let's say I have a method that is annotated with isolation.SERIALIZABLE; and this piece of code is invoked on different nodes on my cluster. Does the isolation level apply only to a single session? Or does it actually effect all the clients? My guess is that it runs across sessions (nodes) but wanted to get some feedback.
addition to the problem statement (as context): Let's say the method is annotated with isolation.SERIALIZABLE and does the following:
Read from table A
Based on the data read do something then update table B
Usually I do these things in a stored-proc but this is happening in java-code. My worry is that the db connection on machine M1 won't be aware of db connection acquired by machine M2 causing concurrency issues. I'm not sure if our connection pool is across machines - even if it were M1 and M2 could use different connections, no?