Most time-based (type 1) UUIDs are created using the current time. I'm migrating a MySQL database to cassandra and would like to create timebased UUIDs for archived items. Can someone provide an example of how type-1 UUIDs are created using past time periods?
+3
A:
All uuids version 1 are a combination of node identifier (MAC address), timestamp and a random seed.
Yes its possible. The process is reversible.
From RFC4122 about the timestamp part of UUID version 1 (section 4.1.4):
"For UUID version 1, this is represented by Coordinated Universal Time (UTC) as a count of 100-nanosecond intervals since 00:00:00.00, 15 October 1582 (the date of Gregorian reform to the Christian calendar)."
Basic algorithm (section 4.2.1) for creating time base UUID (uuid version 1)
Schildmeijer
2010-09-22 06:10:52