The total size of the log record.
Size includes this field. In addition, the size includes padding up to the 8-byte boundary placed after the userlog record.
That is, the next log record header is at a total_len offset from the beginning of the log record header.
Log sequence number.
A sequence number that will be appended to all WAL records, not just userlogs. The following tli and sets can uniquely identify the log.
The timeline ID.
See
For more information about timeline IDs, see “Continuous Archiving and Point-In-Time Recovery (PITR) ” in the“ PostgreSQL Documentation ”. Although not documented in this manual, the timeline also changes when a switchover occurs in a hot standby configuration.
Transaction state.
"c" (for commit) or "a" (for abort).
"true" (integer 1) if this is the last log in the same transaction. Otherwise, it is "false" (integer 0).
The data type of the userlog.
"t" for text type and "b" for bytea type.
Reserved area.
The transaction ID.
The length of content specified to the pgx_emit_userlog function.
The content specified to the pgx_emit_userlog function.
Even type text does not include the terminator of the string.
The character encoding is the database encoding. The database encoding is specified, for example, by the -E option of initdb, or by the WITH ENCODING clause of the CREATE DATABASE statement.
