Updated on 2025-05-29 GMT+08:00

Backup and Restoration

operation_mode

Parameter description: Specifies whether the system enters the backup and restoration mode.

Parameter type: Boolean.

Unit: none

Value range:

  • on: The backup and restoration mode is entered.
  • off: The backup and restoration mode is not entered.

Default value: off

Setting method: This is a SIGHUP parameter. Set it based on instructions provided in Table 1.

Setting suggestion: Retain the default value.

Risks and impacts of improper settings: Changing the value of this parameter may cause unexpected results in the cluster. Retain the default value.

enable_cbm_tracking

Parameter description: Specifies whether to enable the CBM tracing function.

Parameter type: Boolean.

Unit: none

Value range:

  • on: The cbm tracking is enabled.
  • off: The cbm tracking is disabled.

Default value: off

Setting method: This is a SIGHUP parameter. Set it based on instructions provided in Table 1.

Setting suggestion: Enable this parameter to perform full or incremental backup for the cluster by using Roach. Otherwise, the backup will fail.

Risks and impacts of improper settings: If this parameter is disabled, incremental backup will fail.

max_size_for_xlog_retention

Parameter description: Specifies when to forcibly update the backup replication slot or logical replication slot to prevent the disk from being full and the cluster from being read-only because logs cannot be recycled during backup or logical decoding.

Parameter type: integer

Unit: none

Value range: –100 to 2147483647

  • The value 0 indicates that this function is disabled.
  • A negative value indicates that the backup replication slot or logic replication slot is forcibly updated when the disk usage exceeds the threshold and logs are recycled due to blocked backup operation or logical decoding. For example, –80 indicates that the disk usage exceeds 80% of the threshold.
  • A positive value indicates that the backup replication slot or logic replication slot is forcibly updated when the size of stacked logs exceeds the threshold and logs are recycled due to blocked backup operation or logical decoding. For example, 32 indicates that the backup replication slot lags behind the redo location of the current checkpoint by more than 32 log segments. (The size of each log segment is 16 MB.)

Default value: –80

Setting method: This is a SIGHUP parameter. Set it based on instructions provided in Table 1.

Setting suggestion: Set this parameter to a value slightly smaller than that of datastorage_threshold_value_check of the CM Server component to prevent the cluster from entering the read-only state.

Risks and impacts of improper settings: If the value is too small, backup may fail frequently.

max_cbm_retention_time

Parameter description: Specifies the interval at which CBM backup files are forcibly recycled. If CBM files cannot be recycled during backup, the disk will be full and the cluster will be read-only.

Parameter type: integer

Unit: second

Value range: 86400 to 2147483647

Default value: 1209600 (that is, 14 days)

Setting method: This is a SIGHUP parameter. Set it based on instructions provided in Table 1. For example, if this parameter is set to 86400 without a unit, it indicates 86400s. If this parameter is set to 1d, it indicates 1 day. The unit must be s, min, h, or d if required.

Setting suggestion: Set this parameter based on the full backup interval.

Risks and impacts of improper settings: If the value is too small, backup may fail frequently.

enable_standby_move_buckets

Parameter description: Specifies whether a CN is allowed to connect to a standby node to move buckets. This parameter is used only for cluster-level restoration. When restoring the backup set generated during hash bucket redistribution, this parameter is used to establish a connection with the standby DN when the CN sends the buckets migration command to the DN. This parameter takes effect only in this connection. In other cases, the value is off.

Parameter type: Boolean.

Unit: none

Value range:
  • on: The standby node can receive connections.
  • off: The standby node cannot receive connections.

Default value: off

Setting method: This is a BACKEND parameter and cannot be set or reloaded using gs_guc. This parameter does not take effect when written into the configuration file. It is used only as an O&M parameter for data restoration.

Setting suggestion: Retain the default value.

Risks and impacts of improper settings: Changing the value of this parameter may cause unexpected results in the cluster. Retain the default value.