In doing recently when data migration, data tables into the encounter, always prompt string too long or sequence problems. It is easy to locate the problem: the character encoding. View the database character encoding by sql statement: select datname, pg_encoding_to_char (encoding) from pg_database. Was found in a UTF8, and the other is SQL_ASCII, you can change the database character encoding, be consistent.
A first: consistent character encoding and user name of the database.
Second, the transfer of data and then do:
When data migration, by pg_dump command
(1) specify the library specified table pg_dump -d db_name -t table_name -f filename
恢复：psql -h $host -p $port -U $user -W $password -d $database < $backup_file
(2) 压缩 pg_dump -d db_name -t table_name | gzip file_name.gz
恢复：gunzip -c filename.gz | psql dbname 或者 cat filename.gz | gunzip | psql dbname
(3) cut into small files multi-threaded execution pg_dump dbname | split -b 1m - filename
Recovery: cat filename * | psql dbname