Time warp synchronization

The time-warp synchronization method ( German  "time distortion" ) is an optimistic synchronization method, as it is z. B. is used to synchronize different processes in a distributed system.

Time warp synchronization is - like other optimistic synchronization methods - characterized by the following things:

• All processes run without waiting for messages from other processes.
• If messages from the past arrive from other processes, this leads to a loss of causality.
• A process is then able to perform a so-called rollback, i.e. return to the last correct state and continue the calculations taking into account the delayed message from the other process.