Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.databases.postgresql > #537
| From | Dimitri Fontaine <dimitri@2ndQuadrant.fr> |
|---|---|
| Newsgroups | comp.databases.postgresql |
| Subject | Re: PostgreSQL replication using DRBD multi primary + GFS2?? |
| Date | 2014-01-21 16:24 +0100 |
| Organization | A noiseless patient Spider |
| Message-ID | <m2lhy91hpb.fsf@2ndQuadrant.fr> (permalink) |
| References | <cbdd63f8-a168-48d7-a30d-0dd06097afaf@googlegroups.com> <m2k3dvhflg.fsf@2ndQuadrant.fr> <28bcf6e3-95bd-471c-bd81-eac2214aefef@googlegroups.com> |
cancedda@gmail.com writes: > We tried that already but long queries being run on the master break > replication and make running queries on the slave impossible to complete. > > Or that's what I got reading about the error. When using Hot Standby you need to be careful about the setup, which is quite different if you are implementing HA or load balancing. With async replication you always are going to have conflicts to handle, here as the stanbdy is read-only the conflicts are limited to replaying actions from the primary that are incompatible with the queries run on the standby. See the following parameters to help PostgreSQL figure out you're wanting to implement Load Balancing: max_standby_streaming_delay = '20 mins' hot_standby_feedback = on Regards, -- Dimitri Fontaine PostgreSQL DBA, Architecte
Back to comp.databases.postgresql | Previous | Next — Previous in thread | Find similar
PostgreSQL replication using DRBD multi primary + GFS2?? cancedda@gmail.com - 2014-01-19 10:15 -0800
Re: PostgreSQL replication using DRBD multi primary + GFS2?? Dimitri Fontaine <dimitri@2ndQuadrant.fr> - 2014-01-19 21:38 +0100
Re: PostgreSQL replication using DRBD multi primary + GFS2?? cancedda@gmail.com - 2014-01-21 07:01 -0800
Re: PostgreSQL replication using DRBD multi primary + GFS2?? Dimitri Fontaine <dimitri@2ndQuadrant.fr> - 2014-01-21 16:24 +0100
csiph-web