Changes between Version 3 and Version 7 of Ticket #2170

Timestamp:

Oct 5, 2005, 2:45:56 PM (19 years ago)

Author:

Christian Boos

Comment:

Found some relevant discussion about the problem:

• The SQLite API for ​sqlite3_busy_handler describes a similar scenario
  In particular:

Consider a scenario where one process is holding a read lock that it is trying to promote to a reserved lock and a second process is holding a reserved lock that it is trying to promote to an exclusive lock. The first process cannot proceed because it is blocked by the second and the second process cannot proceed because it is blocked by the first. If both processes invoke the busy handlers, neither will make any progress.

• Karl Mierle had a similar issue back in June with Argon, which he describe there ​Locking
• The SQLite author was contacted by the PySQLite author about the former problem, and he ​replied
  • use EXCLUSIVE transactions, but this reduce the amount of concurrency
  • one of the involved threads must ROLLBACK