Race condition with Paged Result searches causes a connection error T3

[mreynolds389]

Issue Description

There is a race condition with Paged Result searches when a new operation comes in while a paged search is finishing. This triggers an invalid time out error and closes the connection.

The problem is that we do not use the "PagedResult lock" when checking the connection's paged result data for a timeout event. This causes the paged result timeout value to change unexpectedly and trigger a false timeout when a new operation arrives.

There are also memory leaks when an error occurs on a paged result search.