Main

CHANGELOG.md

@@ -43,7 +43,7 @@ Pacemaker 1.1.13 using a corosync 2.x stack.
 Changelog since 2.1:
 
 * new: support PostgreSQL 10
-* new: add the maxlag parameter to exclude lagging slaves from promotion, Thomas Reiss
+* new: add the maxlag parameter to exclude lagging secondaries from promotion, Thomas Reiss
 * new: support for multiple pgsqlms resources in the same cluster
 * new: provide comprehensive error messages to crm_mon
 * fix: follow the resource agent man page naming policy and section
@@ -93,7 +93,7 @@ Pacemaker 1.1.13 using a corosync 2.x stack.
 * fix: check on application_name in validate_all
 * change: do not start standby with a master score of 1
 * change: choose the clone to promote when no master score exist
-* new: detect and deal master/slave recovery transition
+* new: detect and deal priamry/secondary recovery transition
 * new: detect and enforce reliability of a switchover
 * new: set next best secondaries base on their lag
 * misc: code cleanup and refactoring
@@ -126,7 +126,7 @@ Release date: 2016-04-27
 * fix: OCF tests when PostgreSQL does not listen in /tmp
 * change: do not update score outside of a monitor action (gh #18)
 * new: add parameter 'start_opts', usefull for debian and derivated (gh #11)
-* new: add specific timeout for master and slave roles in meta-data (gh #14)
+* new: add specific timeout for primary and secondary roles in meta-data (gh #14)
 * new: add debian packaging related files
 
 
@@ -136,4 +136,3 @@ Release date: 2016-04-27
 Release date: 2016-03-02
 
 * First public release
-

README.md

@@ -24,11 +24,11 @@ yet powerful.
 
 Once your PostgreSQL cluster built using internal streaming replication, PAF is
 able to expose to Pacemaker what is the current status of the PostgreSQL
-instance on each node: master, slave, stopped, catching up, etc. Should a
-failure occurs on the master, Pacemaker will try to recover it by default.
-Should the failure be non-recoverable, PAF allows the slaves to be able to
-elect the best of them (the closest one to the old master) and promote it as
-the new master. All of this thanks to the robust, feature-full and most
+instance on each node: primary, secondary, stopped, etc. Should a
+failure occurs on the primary, Pacemaker will try to recover it by default.
+Should the failure be non-recoverable, PAF allows the secondaries to be able to
+elect the best of them (the closest one to the old primary) and promote it as
+the new primary. All of this thanks to the robust, feature-full and most
 importantly experienced project: Pacemaker.
 
 For information about how to install this agent, see `INSTALL.md`.
@@ -40,10 +40,10 @@ CentOS 6 and 7 in various scenario.
 
 PAF has been written to give to the administrator the maximum control
 over their PostgreSQL configuration and architecture. Thus, you are 100%
-responsible for the master/slave creations and their setup. The agent
+responsible for the primary/secondary creations and their setup. The agent
 will NOT edit your setup. It only requires you to follow these pre-requisites:
 
-  * slave __must__ be in hot_standby (accept read-only connections) ;
+  * secondary __must__ be in hot_standby (accept read-only connections) ;
   * the following parameters __must__ be configured in the appropriate place :
     * `standby_mode = on` (for PostgreSQL 11 and before)
     * `recovery_target_timeline = 'latest'`
@@ -66,7 +66,7 @@ can set:
   * `pghost`: the socket directory or IP address to use to connect to the
     local instance (default: `/tmp` or `/var/run/postgresql` for DEBIAN)
   * `pgport`:  the port to connect to the local instance (default: `5432`)
-  * `recovery_template`: __only__ for PostgreSQL 11 and before. The local 
+  * `recovery_template`: __only__ for PostgreSQL 11 and before. The local
     template that will be copied as the `PGDATA/recovery.conf` file. This
     file must not exist on any node for PostgreSQL 12 and after.
     (default: `$PGDATA/recovery.conf.pcmk`)
@@ -76,11 +76,10 @@ can set:
     `-c config_file=/etc/postgresql/9.3/main/postgresql.conf`
   * `system_user`: the system owner of your instance's process (default:
     `postgres`)
-  * `maxlag`: maximum lag allowed on a standby before we set a negative master
-    score on it. The calculation is based on the difference between the current
-    xlog location on the master and the write location on the standby.
-    (default: 0, which disables this feature)
+  * `maxlag`: maximum lag allowed on a standby before we set a negative
+    promotion score on it. The calculation is based on the difference between
+    the current xlog location on the primary and the write location on the
+    standby. (default: 0, which disables this feature)
 
 For a demonstration about how to setup a cluster, see
 [http://clusterlabs.github.io/PAF/documentation.html](http://clusterlabs.github.io/PAF/documentation.html).
-

debian/changelog

@@ -108,53 +108,28 @@ resource-agents-paf (2.0.0-1) unstable; urgency=low
 resource-agents-paf (2.0~rc1-1) unstable; urgency=low
 
   * 2.0_rc1 first release candidate
+  * see CHANGELOG.md
 
  -- Jehan-Guillaume (ioguix) de Rorthais <jgdr@dalibo.com>  Wed, 3 Aug 2016 18:17:35 +0200
 
 resource-agents-paf (2.0~beta2-1) unstable; urgency=low
 
   * 2.0_beta2 beta release
-  * fix: bug in switchover with 9.5 and 9.6
+  * see CHANGELOG.md
 
  -- Jehan-Guillaume (ioguix) de Rorthais <jgdr@dalibo.com>  Fri, 1 Jul 2016 16:35:35 +0200
 
 resource-agents-paf (2.0~beta1-1) unstable; urgency=low
 
   * 2.0_beta1 beta release
-  * fix: do not use crm_node --partition to discover resources
-  * fix: unknown argument --query when calling crm_master
-  * fix: perl warning when master score has never been set on the master
-  * fix: remove wrong info message during post-promote notify
-  * fix: race condition when setting attributes during actions
-  * fix: bug where pgport and pghost where ignored in _query
-  * fix: use same role name than the system_user to connect
-  * fix: wrap crm_master calls in sub to make them synchronous
-  * fix: fixed a bug related to setgid in _runas
-  * fix: check on application_name in validate_all
-  * change: do not start standby with a master score of 1
-  * change: choose the clone to promote when no master score exist
-  * new: detect and deal master/slave recovery transition
-  * new: detect and enforce reliability of a switchover
-  * new: set next best secondaries base on their lag
-  * misc: code cleanup and refactoring
-  * misc: various log messages cleanup and enhancement
+  * see CHANGELOG.md
 
  -- Jehan-Guillaume (ioguix) de Rorthais <jgdr@dalibo.com>  Wed, 15 Jun 2016 23:19:58 +0200
 
 resource-agents-paf (1.0.1-1) unstable; urgency=low
 
   * 1.0.1 minor release
-  * fix: forbid the master to decrease its own score (gh #19)
-  * fix: bad LSN decimal converstion (gh #20)
-  * fix: support PostgreSQL 9.5 controldata output (gh #12)
-  * fix: set group id of given system_user before executing commands (gh #11)
-  * fix: use long argument of external commands when possible
-  * fix: bad header leading to wrong manpage section
-  * fix: OCF tests when PostgreSQL does not listen in /tmp
-  * change: do not update score outside of a monitor action (gh #18)
-  * new: add parameter 'start_opts', usefull for debian and derivated (gh #11)
-  * new: add specific timeout for master and slave roles in meta-data (gh #14)
-  * new: add debian packaging related files
+  * see CHANGELOG.md
 
  -- Jehan-Guillaume (ioguix) de Rorthais <jgdr@dalibo.com>  Wed, 27 Apr 2016 13:22:50 +0200
 

debian/control

@@ -17,4 +17,4 @@ Description: PostgreSQL resource agent for Pacemaker
  the Pacemaker administration and the PostgreSQL one, to keep things simple,
  documented and yet powerful.
  .
- It only supports multi-state (Master-Slave) PostgreSQL clusters.
+ It only supports multi-state (Primary-Secondary) PostgreSQL clusters.

extra/vagrant/2nodes-qdevice-vip/provision/pgsql.bash

@@ -83,14 +83,14 @@ if [ "$NODENAME" == "$PRIM_NODE" ]; then
 
     chown -R postgres:postgres "$PGDATA"
 
-    # create master ip
+    # create primary ip
     ip -o addr show to "${PRIM_IP}" | if ! grep -q "${PRIM_IP}"
     then
         DEV=$(ip route show to "${PRIM_IP}/24"|grep -Eo 'dev \w+')
         ip addr add "${PRIM_IP}/24" dev "${DEV/dev }"
     fi
 
-    # restart master pgsql
+    # restart primary pgsql
     systemctl --quiet start "postgresql-${PGVER}"
 
     exit

extra/vagrant/3nodes-haproxy/provision/pacemaker.bash

@@ -85,7 +85,7 @@ PGSQLD_RSC_OPTS=(
 # NB: pcs 0.10.2 doesn't support to set the id of the clone XML node
 # the id is built from the rsc id to clone using "<rsc-id>-clone"
 # As a matter of cohesion and code simplicity, we use the same
-# convention to create the master resource with pcs 0.9.x for
+# convention to create the primary resource with pcs 0.9.x for
 # Pacemaker 1.1
 if [ "$PCMK_VER" -ge 2 ]; then
     PGSQLD_RSC_OPTS+=( "promotable" "notify=true" )

extra/vagrant/3nodes-haproxy/provision/pgsql.bash

@@ -83,7 +83,7 @@ if [ "$NODENAME" == "$PRIM_NODE" ]; then
 
     chown -R postgres:postgres "$PGDATA"
 
-    # restart master pgsql
+    # restart primary pgsql
     systemctl --quiet start "postgresql-${PGVER}"
 
     exit

extra/vagrant/3nodes-vip/provision/pgsql.bash

@@ -83,7 +83,7 @@ if [ "$NODENAME" == "$PRIM_NODE" ]; then
 
     chown -R postgres:postgres "$PGDATA"
 
-    # create master ip
+    # create primary ip
     ip -o addr show to "${PRIM_IP}" | if ! grep -q "${PRIM_IP}"
     then
         DEV=$(ip route show to "${PRIM_IP}/24"|grep -Eo 'dev \w+')

lib/OCF_Functions.pm

@@ -12,7 +12,7 @@ OCF_Functions - helper subroutines for OCF agent
 
   use FindBin;
   use lib "$FindBin::RealBin/../../lib/heartbeat/";
-  
+
   use OCF_Functions;
 
 =head1 DESCRIPTION
@@ -257,7 +257,7 @@ sub ha_debug {
         else {
             printf STDERR "%s\n", join ' ', @ARG;
         }
-        
+
         return 0;
     }
 
@@ -375,9 +375,9 @@ sub ocf_is_clone {
         and $ENV{'OCF_RESKEY_CRM_meta_clone_max'} > 0 );
 }
 
-# returns true if the resource is configured as a multistate
-# (master/slave) resource. This is defined as a resource where the
-# master-max meta attribute is present, and set to greater than zero.
+# returns true if the resource is promotable.
+# This is defined as a resource where the master-max meta attribute is present,
+# and set to greater than zero.
 sub ocf_is_ms {
     return ( defined $ENV{'OCF_RESKEY_CRM_meta_master_max'}
         and  $ENV{'OCF_RESKEY_CRM_meta_master_max'} > 0 );
@@ -394,7 +394,7 @@ sub ocf_is_ver {
 
 sub ocf_ver2num {
     my $v = 0;
-    
+
     $v = $v * 1000 + $1 while $ARG[0] =~ /(\d+)/g;
 
     return $v;
@@ -431,7 +431,7 @@ sub ocf_version_cmp {
     my $v1_level;
     my $v2_level;
     my $level_diff;
-    
+
     return 3 unless ocf_is_ver( $v1 );
     return 3 unless ocf_is_ver( $v2 );