When the node gets synchronized up to the minimum recovery

lag, all the cluster nodes get locked for write transactions to allow the

recovery process to finish. By default, the minimum recovery

lag is100KB. You can change this value in the

lag is10MB. You can change this value in the

<varname>multimaster.min_recovery_lag</varname> variable.

</para>

</listitem>

<para>

<varname>multimaster.min_recovery_lag</varname> &mdash; sets the

minimal WAL lag between the node to be restored and the current cluster state.

By default, <varname>multimaster.min_recovery_lag</varname> is set to100KB.

By default, <varname>multimaster.min_recovery_lag</varname> is set to10MB.

When the disconnected node is fast-forwarded up to the

<varname>multimaster.min_recovery_lag</varname> threshold,

<filename>multimaster</filename> stops all new commits to the

</itemizedlist>

<para>

By default, <varname>multimaster.max_recovery_lag</varname> is

set to100MB. Setting

set to1GB. Setting

<varname>multimaster.max_recovery_lag</varname> to a larger

value increases the timeframe for automatic recovery, but

requires more disk space for WAL collection.

<varlistentry><term><varname>multimaster.min_recovery_lag</varname><indexterm><primary><varname>multimaster.min_recovery_lag</varname></primary></indexterm></term><listitem><para>

Minimal WAL lag

between the node to be restored and the current cluster state, in

bytes. When this threshold is reached during node recovery, the

kB. When this threshold is reached during node recovery, the

cluster is locked for write transactions until the recovery is

complete. </para><para>Default:100000

complete. </para><para>Default:10MB

</para></listitem></varlistentry>

<varlistentry><term><varname>multimaster.max_recovery_lag</varname><indexterm><primary><varname>multimaster.max_recovery_lag</varname></primary></indexterm></term><listitem><para>

Maximal WAL lag

size, inbytes. When a node is disconnected from the cluster, other

size, inkB. When a node is disconnected from the cluster, other

nodes copy WALs for all new transactions into the replication slot of

this node. Upon reaching the

<varname>multimaster.max_recovery_lag</varname> value, the

the data from one of the alive nodes using

<application>pg_basebackup</application> or a similar tool. If you set this

variable to zero, replication slot will not be dropped. </para><para>Default:

</para></listitem></varlistentry>

<varlistentry id="ignore-tables-without-pk"><term><varname>multimaster.ignore_tables_without_pk</varname><indexterm><primary><varname>multimaster.ignore_tables_without_pk</varname></primary></indexterm></term><listitem><para>

Boolean.

</indexterm>

</term>

<listitem>

<para>The maximal size of transaction, inMB. When this threshold is reached, the transaction is written to the disk.

<para>The maximal size of transaction, inkB. When this threshold is reached, the transaction is written to the disk.

</para>

<para>Default:100

<para>Default:100MB

</para>

</listitem>

</varlistentry>