Re: rsync and streaming replication
От | Cédric Villemain |
---|---|
Тема | Re: rsync and streaming replication |
Дата | |
Msg-id | CAF6yO=3fjuokv9J80oW9oQWF9YdYxMcH_gSyfJ5KCKwiA5V_5w@mail.gmail.com обсуждение исходный текст |
Ответ на | Re: rsync and streaming replication (Jean-Armel Luce <jaluce06@gmail.com>) |
Ответы |
Re: rsync and streaming replication
|
Список | pgsql-admin |
2011/11/16 Jean-Armel Luce <jaluce06@gmail.com>: > Hi, > > Today I tried to promote a slave as master using rsync --checksum (without > doing vacuum freeze) instead of rsync --all. > > It takes only 30 minutes with rsync --checksum. Only a few tables are > rsynced. Most of the time is consumed by checksum. > With rsync --all, it takes 1h40 min. > > So, rsync --checksum looks better than rsync --all rsync --all does not exist, you mean --archive I bet (the longopt for -a, a shorthand for the options below plus -Dt for device/special and time). Also you probably want to keep --owner --group --perms --recursive (or you do file by file with a find or something like that ?) -- Cédric Villemain +33 (0)6 20 30 22 52 http://2ndQuadrant.fr/ PostgreSQL: Support 24x7 - Développement, Expertise et Formation
В списке pgsql-admin по дате отправления: