Skip to content

Latest commit

 

History

History
777 lines (594 loc) · 49.8 KB

user_manual-EN.md

File metadata and controls

777 lines (594 loc) · 49.8 KB

User Guide

Create a PVC

The created PVC is used as the storage for the database cluster.

Field Description:

Field Description
IP IP address of the machine where the storage management pod is located, e.g., 10.0.0.77.
name Name of the PVC to be created, e.g., pvc-32ze341nncwlczm47bsre
volume_id wwid of the shared storage, e.g., 32ze341nncwlczm47bsre. This disk should support the SCSI or NVMe protocol.

Example:

curl -X POST "http://${IP}:2002/pvcs" -H "accept: application/json" -H "Content-Type: application/json" -d "{ \"name\": \"${name}\", \"namespace\": \"default\", \"need_format\": true, \"volume_id\": \"${volume_id}\", \"volume_type\": \"lun\"}"

If the following information is returned, it indicates that the PVC is created successfully.

{"workflow_id":"7cc5191d-803f-4ea5-8eb1-92e5437a52e9"}

After the PVC is created, you can view the created PVC using the following command in Kubernetes:

kubectl get pvc

Create an Instance Cluster

Create an instance cluster. You can access your database by accessing this cluster.

Steps:

  1. Configure the YAML file of the instance cluster according to your actual needs and execute the command kubectl to make it take effect.
kubectl apply -f your-cluster.yaml

Field Description:

Field Field Description
operatorName The operator name you are using. By default it is polar-mpd and you can configure it using the parameter --filter-operator-name in the operator startup parameters.
dbClusterType The storage type is shared storage.
followerNum Number of ro nodes that need to be created.
classInfo.className Name of the instance's specification, which corresponds to the classKey in instance_level_config you have configured, e.g., polar.o.x4.medium. You can search all.yaml to find all configurations.
versionCfg.versionName The suffix of the corresponding image configuration configmap postgresql-1-0-minor-version-info-rwo-${versionCfg.versionName}. The image-open is the default configuration example in all.yaml.
netCfg.engineStartPort Port used to create database services.
shareStore.sharePvcName Name of the created PVC.
shareStore.volumeId The wwid of the corresponding shared disk of your PVC. This disk MUST support SCSI or NVMe protocol.
shareStore.diskQuota The disk size that can be used by your database engine, unit: MB. As PFS uses 10 GB as a unit, only a multiple of 10 GB is allowed for increasing the capacity each time.

Example:

apiVersion: mpd.polardb.aliyun.com/v1
kind: MPDCluster
metadata:
  name: mpdcluster-sample-2
  namespace: default
spec:
  operatorName: polar-mpd
  dbClusterType: share
  followerNum: 1
  classInfo:
    className: polar.o.x4.medium
  classInfoModifyTo:
    className: ""
  versionCfg:
    versionName: image-open
  netCfg:
    engineStartPort: 5780
  shareStore:
    drive: "pvc"
    sharePvcNamespace: "default"
    sharePvcName: "pvc-32ze341nncwlczm47bsre"
    diskQuota: "300000"
    volumeId: "32ze341nncwlczm47bsre"
    volumeType: "multipath"

Verification result:

  1. clusterStatus: Running indicates that the cluster is created successfully.
[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: Running

Change Specifications

Change the specifications of your instance.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Modify the field classInfoModifyTo.className to the name of the specification you want to change the current one to.

Field Description:

Field Description
classInfoModifyTo.className Name of the specification after change, which corresponds to classKey in instance_level_config you have configured.

Example:

spec:   
  classInfoModifyTo:
    className: polar.o.x4.large

Verify the status:

  • During the process of changing specifications, the clusterStatus will be changed to ModifyClass.
[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com mpdcluster-sample-1 -o yaml | grep clusterStatus
  clusterStatus: ModifyClass
  • After changing specifications is completed, the clusterStatus will be changed to Running again.
[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com mpdcluster-sample-1 -o yaml | grep clusterStatus
  clusterStatus: Running

Switch between the RW Node and RO Nodes

Switch a ro node to the rw node in your database cluster.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.switchRw The ID of the ro node to be switched to the rw node. The ID is listed in status.dbInstanceStatus.

Example:

metadata:
  annotations:
    switchRw: "7614"

Verify the status:

Switching

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: SwitchRw

Modify Engine Parameters

Modify the engine parameters of the created instance.

Steps:

  1. View the parameters which are saved in the configmap named ${your-cluster-name}-user-params.
kubectl edit cm your-cluster-name-user-params
  1. Edit the parameters you want to modify. For the value range of each parameter, refer to the Appendix. For example:
checkpoint_timeout: '{"name":"checkpoint_timeout","value":"300","updateTime":"2021-09-29T07:28:59Z","inoperative_value":""}'

  1. You can modify the value of the field value to the one you need and change the value of the field updateTime to the current time. As updating parameters requires verifying updateTime, remember you MUST update the value of the field updateTime.

  2. After modifying the above configmap, edit your cluster and set the parameter modification operation.

kubectl edit mpdcluster your-cluster-name

Add the following field whose description is shown in the table below.

Field Field Description
metadata.annotations.flushParams If this field is set to true, it indicates that modifying parameters is required.

Example:

metadata:
  annotations:
    flushParams: "true"

Restart an Instance

You can restart a specific node in the database cluster.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.restartIns The ID of the database instance you want to restart. The ID is listed in status.dbInstanceStatus as a number.

Example:

metadata:
  annotations:
    restartIns: "7503"

Restart a Cluster

Restart the entire database cluster.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.restartCluster If this field is set to true, it indicates that restarting the cluster is required.

Example:

metadata:
  annotations:
    restartCluster: "true"

Update the Image Version

Modify configmap postgresql-1-0-minor-version-info-rwo-image-open.

kubectl -n kube-system edit configmap postgresql-1-0-minor-version-info-rwo-image-open

Update the image version that you need to modify as the example below:

apiVersion: v1
data:
  name: image-open
  pfsdImage: polardb/pfsd:1.2.41-20211018
  pfsdToolImage: polardb/pfsd_tool:1.2.41-20211018
  pgClusterManagerImage: polardb/polardb-cluster-manager:latest
  pgEngineImage: polardb/polardb_pg_engine_release:11beta2.20210910.d558886c.20211018195123
  pgManagerImage: polardb/polardb_pg_manager:20211018195123.9ae43314
kind: ConfigMap
metadata:
  labels:
    configtype: minor_version_info
    dbClusterMode: WriteReadMore
    dbType: PostgreSQL
    dbVersion: "1.0"
  name: postgresql-1-0-minor-version-info-rwo-image-open
  namespace: kube-system

Field Description:

Field Field Description
pgEngineImage Image of the database engine.
pgManagerImage Image of the manager
pfsdImage Image of the pfsd
pfsdToolImage Image of the pfsd tool

Upgrade the Minor Version

Upgrade the image version used by your database cluster.

Steps:

  1. Create a new configuration for the image, modify the configuration parameters to those for the image version you need to upgrade to, and use the command kubectl to make the configuration take effect.

Field Description:

Field Field Description
pgEngineImage Image of the database engine
pgManagerImage Image of the manager
pfsdImage Image of the pfsd
pfsdToolImage Image of the pfsd tool 
apiVersion: v1
data:
  name: image-open
  pfsdImage: polardb/pfsd:1.2.41-20211018
  pfsdToolImage: polardb/pfsd_tool:1.2.41-20211018
  pgClusterManagerImage: polardb/polardb-cluster-manager:latest
  pgEngineImage: polardb/polardb_pg_engine_release:11beta2.20210910.d558886c.20211018195123
  pgManagerImage: polardb/polardb_pg_manager:20211018195123.9ae43314
kind: ConfigMap
metadata:
  labels:
    configtype: minor_version_info
    dbClusterMode: WriteReadMore
    dbType: PostgreSQL
    dbVersion: "1.0"
  name: postgresql-1-0-minor-version-info-rwo-image-open
  namespace: kube-system
  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Modify the field versionCfgModifyTo.versionName to the name of the specification you want to change the current one to. Add the following field whose description is shown in the table below.
Field Field Description
versionCfgModifyTo.versionName The name of the specification you want to change the current one to, e.g., "image-open".

Example:

spec: 
  versionCfgModifyTo:
    versionName: your-new-config-name

Recreate Completely

Recreate the database cluster.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.forceRebuild If this field is set to true, it indicates that recreating the database cluster is required.

Example:

metadata:
  annotations:
    forceRebuild: "true"

Check the status:

Recreating:

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: Rebuild

Recreated:

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: Running

Add a RO Node

Add a ro (read-only) node.

Note: Make sure you have spare machines. Otherwise, a port conflict will occur.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
followerNum Total number of ro nodes after adding.

Example:

spec:
  followerNum: 2

Migrate the RW Node

Migrate the rw (read-write) node to another machine.

Note: The node can only be migrated to a node without any cluster instance. Otherwise, the port conflicts will occur.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.migrate The format is ${insId}|${targetNode} (e.g., 7601|dbm-02), in which the insid is the ID of the current rw node and the targetNode is the name of the destination machine. The targetNode can be obtained using the command kubectl get node.

Example:

metadata:
  annotations:
    migrate: 7601|dbm-02

Verify the status

Migrating:

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: MigrateRw

Migrated:

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: Running

Migrate a RO Node

Migrate a ro (read-only) node to another machine.

Note: The node can only be migrated to a node without any cluster instance. Otherwise, the port conflicts will occur.

Steps:

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.migrate The format is ${insId}|${targetNode} (e.g., 7601|dbm-02), in which the insid is the ID of the ro node to be migrated and the targetNode is the name of the destination machine. The targetNode can be obtained using the command kubectl get node.

Example:

metadata:
  annotations:
    migrate: 7601|dbm-02

Verify the status:

Migrating:

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: MigrateRo

Migrated:

[root@dbm-01 ~]# kubectl get mpdclusters.mpd.polardb.aliyun.com your-cluster-name -o yaml | grep clusterStatus
  clusterStatus: Running

Increase Storage Capacity

Increase the storage capacity used by your database.

Before that, you need to increase the disk capacity.

  1. Modify the created cluster.
kubectl edit mpdcluster your-cluster-name
  1. Add the following field whose description is shown in the table below.
Field Field Description
metadata.annotations.extendStorage If this field is set to true, it indicates that increasing the storage capacity is required.

Modify the value of the following field to the disk size you need after increasing. The field is described as follows:

Field Field Description
spec.shareStore.diskQuota The disk size that can be used by your database engine, unit: MB. As PFS uses 10 GB as a unit, only a multiple of 10 GB is allowed for increasing the capacity each time.

Example:

metadata:
  annotations:
    extendStorage: "true"

Retry after Failure

This operation is suitable for the situation in which the operation failed and the status is changed to Interrupt.

Change the value of interrupt.recover from F to T.

metadata:
  annotations:
    interrupt.recover: F

Delete the Instance Cluster

Perform this operation with caution.

kubectl delete mpdcluster your-cluster-name

Appendix A

The engine parameters are described as follows.

Parameter Name Default Value Restarting Required or Not Value Range
archive_mode on Y [always|on|off]
autovacuum_vacuum_cost_delay 0 N [-1-100]
autovacuum_vacuum_cost_limit 10000 N [-1-10000]
auto_explain.log_analyze off N [on|off]
auto_explain.log_buffers off N [on|off]
auto_explain.log_format text N [text|xml|json|yaml]
auto_explain.log_min_duration 5000 N [-1-2147483647]
auto_explain.log_nested_statements off N [on|off]
auto_explain.log_timing on N [on|off]
auto_explain.log_triggers off N [on|off]
auto_explain.log_verbose off N [on|off]
auto_explain.sample_rate 1 N [0-1]
checkpoint_completion_target 0.9 N [0-1]
checkpoint_timeout 300 N [1-86400]
default_transaction_deferrable off N [on|off]
default_with_oids off N [on|off]
default_with_rowids off N [on|off]
enable_partitionwise_aggregate on N [on|off]
enable_partitionwise_join on N [on|off]
extra_float_digits 0 N [-15-3]
idle_in_transaction_session_timeout 3600000 N ^(0|[1-9]\d{3,8}|1\d{9}|2000000000)$
jit off N [on|off]
lock_timeout 0 N ^(0|[1-9]\d{3,8}|1\d{9}|2000000000)$
log_min_duration_statement -1 N [-1-2147483647]
log_statement all N [none|ddl|mod|all]
log_temp_files 100000 N [-1-2147483647]
log_timezone 'PRC' N ^'(((UTC)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|((GMT)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|CST6CDT|Poland|Kwajalein|MST|NZ|Universal|Libya|Turkey|EST5EDT|Greenwich|NZ-CHAT|MET|Portugal|GMT-0|CET|Eire|PST8PDT|Jamaica|GMT|Zulu|Japan|ROC|GB-Eire|ROK|Navajo|Singapore|posixrules|GB|EST|GMT0|Hongkong|PRC|Iran|MST7MDT|WET|W-SU|UCT|Cuba|Egypt|EET|Israel|UTC|HST|Iceland)'$
max_parallel_maintenance_workers 2 N [0-1024]
max_parallel_workers 2 N [0-512]
max_parallel_workers_per_gather 2 N [0-512]
min_parallel_index_scan_size 64 N [0-715827882]
min_parallel_table_scan_size 1024 N [0-715827882]
old_snapshot_threshold -1 Y [-1-86400]
polar_comp_dynatune 0 Y [0-100]
polar_comp_dynatune_profile oltp Y [oltp|reporting|mixed]
polar_comp_enable_pruning on N [on|off]
polar_comp_redwood_date on N [on|off]
polar_comp_redwood_greatest_least on N [on|off]
polar_comp_redwood_raw_names on N [on|off]
polar_comp_redwood_strings on N [on|off]
polar_comp_stmt_level_tx on N [on|off]
dbms_job.database_name 'postgres' Y ^'\w+'$
statement_timeout 0 N ^(0|[1-9]\d{3,8}|1\d{9}|2000000000)$
temp_file_limit 524288000 N [-1-1048576000]
timezone 'PRC' N ^'(((UTC)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|((GMT)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|CST6CDT|Poland|Kwajalein|MST|NZ|Universal|Libya|Turkey|EST5EDT|Greenwich|NZ-CHAT|MET|Portugal|GMT-0|CET|Eire|PST8PDT|Jamaica|GMT|Zulu|Japan|ROC|GB-Eire|ROK|Navajo|Singapore|posixrules|GB|EST|GMT0|Hongkong|PRC|Iran|MST7MDT|WET|W-SU|UCT|Cuba|Egypt|EET|Israel|UTC|HST|Iceland)'$
track_commit_timestamp off Y [on|off]
vacuum_defer_cleanup_age 0 N [0-1000000]
wal_level logical Y [replica|logical]
work_mem 4096 N [4096-524288]
polar_empty_string_is_null_enable on N [on|off]
polar_enable_varchar2_length_with_byte on N [on|off]
polar_enable_base64_decode on N [on|off]
polar_enable_nls_date_format on N [on|off]
polar_enable_rowid on N [on|off]
postgres_fdw.polar_connection_check off N [on|off]
polar_comp_custom_plan_tries 5 N [-1-100]
dblink.polar_auto_port_mapping off N [on|off]
dblink.polar_connection_check off N [on|off]
pg_stat_statements.enable_superuser_track on N [on|off]
cron.polar_allow_superuser_task on Y [on|off]
polar_stat_sql.enable_qps_monitor on N [on|off]
polar_enable_audit_log_bind_sql_parameter off N [on|off]
polar_enable_audit_log_bind_sql_parameter_new off N [on|off]
polar_enable_replica_use_smgr_cache on Y [off|on]
idle_session_timeout 0 N [0-2147483647]
polar_resource_group.total_mem_limit_rate 95 N [50-100]
polar_resource_group.total_mem_limit_remain_size 524288 N [131072-2097151]

Appendix B

The engine parameters are described as follows.

Parameter Name Default Value Restarting Required or Not Value Range
archive_mode on Y [always|on|off]
autovacuum_vacuum_cost_delay 0 N [-1-100]
autovacuum_vacuum_cost_limit 10000 N [-1-10000]
auto_explain.log_analyze off N [on|off]
auto_explain.log_buffers off N [on|off]
auto_explain.log_format text N [text|xml|json|yaml]
auto_explain.log_min_duration 5000 N [-1-2147483647]
auto_explain.log_nested_statements off N [on|off]
auto_explain.log_timing on N [on|off]
auto_explain.log_triggers off N [on|off]
auto_explain.log_verbose off N [on|off]
auto_explain.sample_rate 1 N [0-1]
checkpoint_completion_target 0.9 N [0-1]
checkpoint_timeout 300 N [1-86400]
default_transaction_deferrable off N [on|off]
default_with_oids off N [on|off]
default_with_rowids off N [on|off]
enable_partitionwise_aggregate on N [on|off]
enable_partitionwise_join on N [on|off]
extra_float_digits 0 N [-15-3]
idle_in_transaction_session_timeout 3600000 N ^(0|[1-9]\d{3,8}|1\d{9}|2000000000)$
jit off N [on|off]
lock_timeout 0 N ^(0|[1-9]\d{3,8}|1\d{9}|2000000000)$
log_min_duration_statement -1 N [-1-2147483647]
log_statement all N [none|ddl|mod|all]
log_temp_files 100000 N [-1-2147483647]
log_timezone 'PRC' N ^'(((UTC)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|((GMT)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|CST6CDT|Poland|Kwajalein|MST|NZ|Universal|Libya|Turkey|EST5EDT|Greenwich|NZ-CHAT|MET|Portugal|GMT-0|CET|Eire|PST8PDT|Jamaica|GMT|Zulu|Japan|ROC|GB-Eire|ROK|Navajo|Singapore|posixrules|GB|EST|GMT0|Hongkong|PRC|Iran|MST7MDT|WET|W-SU|UCT|Cuba|Egypt|EET|Israel|UTC|HST|Iceland)'$
max_parallel_maintenance_workers 2 N [0-1024]
max_parallel_workers 2 N [0-512]
max_parallel_workers_per_gather 2 N [0-512]
min_parallel_index_scan_size 64 N [0-715827882]
min_parallel_table_scan_size 1024 N [0-715827882]
old_snapshot_threshold -1 Y [-1-86400]
polar_comp_dynatune 0 Y [0-100]
polar_comp_dynatune_profile oltp Y [oltp|reporting|mixed]
polar_comp_enable_pruning on N [on|off]
polar_comp_redwood_date on N [on|off]
polar_comp_redwood_greatest_least on N [on|off]
polar_comp_redwood_raw_names on N [on|off]
polar_comp_redwood_strings on N [on|off]
polar_comp_stmt_level_tx on N [on|off]
dbms_job.database_name 'postgres' Y ^'\w+'$
statement_timeout 0 N ^(0|[1-9]\d{3,8}|1\d{9}|2000000000)$
temp_file_limit 524288000 N [-1-1048576000]
timezone 'PRC' N ^'(((UTC)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|((GMT)(-){0,1}(\d|[1-9]\d|1([0-5]\d|6[0-7])))|CST6CDT|Poland|Kwajalein|MST|NZ|Universal|Libya|Turkey|EST5EDT|Greenwich|NZ-CHAT|MET|Portugal|GMT-0|CET|Eire|PST8PDT|Jamaica|GMT|Zulu|Japan|ROC|GB-Eire|ROK|Navajo|Singapore|posixrules|GB|EST|GMT0|Hongkong|PRC|Iran|MST7MDT|WET|W-SU|UCT|Cuba|Egypt|EET|Israel|UTC|HST|Iceland)'$
track_commit_timestamp off Y [on|off]
vacuum_defer_cleanup_age 0 N [0-1000000]
wal_level logical Y [replica|logical]
work_mem 4096 N [4096-524288]
polar_empty_string_is_null_enable on N [on|off]
polar_enable_varchar2_length_with_byte on N [on|off]
polar_enable_base64_decode on N [on|off]
polar_enable_nls_date_format on N [on|off]
polar_enable_rowid on N [on|off]
postgres_fdw.polar_connection_check off N [on|off]
polar_comp_custom_plan_tries 5 N [-1-100]
dblink.polar_auto_port_mapping off N [on|off]
dblink.polar_connection_check off N [on|off]
pg_stat_statements.enable_superuser_track on N [on|off]
cron.polar_allow_superuser_task on Y [on|off]
polar_stat_sql.enable_qps_monitor on N [on|off]
polar_enable_audit_log_bind_sql_parameter off N [on|off]
polar_enable_audit_log_bind_sql_parameter_new off N [on|off]
polar_enable_replica_use_smgr_cache on Y [off|on]
idle_session_timeout 0 N [0-2147483647]
polar_resource_group.total_mem_limit_rate 95 N [50-100]
polar_resource_group.total_mem_limit_remain_size 524288 N [131072-2097151]

The following table lists high-risk parameters. Confirm the impact first and modify them with caution.

Parameter Name Default Value Restarting Required or Not Value Range
autovacuum_max_workers 5 Y [1-262143]
autovacuum_work_mem 200MB N [-1-2147483647]
enable_hashagg off N [on|off]
log_connections off N [on|off]
log_disconnections off N [on|off]
max_standby_streaming_delay 30000 N [-1-2147483647]
polar_bgwriter_batch_size_flushlist 100 Y [1-10000]
polar_bgwriter_max_batch_size 5000 N [0-1073741823]
polar_bgwriter_sleep_lsn_lag 100 N [0-1000]
polar_buffer_copy_lsn_lag_with_cons_lsn 100 N [1-1000]
polar_buffer_copy_min_modified_count 5 N [0-100]
polar_bulk_extend_size 512 N [0-1073741823]
polar_check_checkpoint_legal_interval 10 N [1-3600]
polar_clog_slot_size 512 Y [128-8192]
polar_comp_early_lock_release off N [on|off]
polar_copy_buffers 16384 Y [128-1073741823]
polar_enable_connectby_multiexpr on N [on|off]
polar_enable_physical_repl_non_super_wal_snd off N [on|off]
polar_enable_show_polar_comp_version on N [on|off]
polar_enable_syslog_pipe_buffer off N [on|off]
polar_global_temp_table_debug off N [on|off]
polar_max_log_files 20 N [-1-2147483647]
polar_max_super_conns 100 N [-1-262143]
polar_num_active_global_temp_table 1000 Y [0-1000000]
polar_parallel_bgwriter_check_interval 10 N [1-600]
polar_parallel_bgwriter_delay 10 N [1-10000]
polar_parallel_bgwriter_workers 5 N [0-8]
polar_parallel_new_bgwriter_threshold_time 10 N [1-3600]
polar_read_ahead_xlog_num 200 Y [0-200]
polar_redo_hashtable_size 131072 Y [16-1073741823]
polar_ring_buffer_vacuum 128 N [10-1000]
polar_spl_savepoint_enable on N [on|off]
polar_temp_relation_file_in_shared_storage off Y [on|off]
polar_vfs.libmm_num_partition 32 Y [4-128]
polar_vfs.libmm_size 67108864 Y [67108864-1073741824]
polar_enable_fullpage_snapshot off N [on|off]
polar_enable_default_polar_comp_guc on N [on|off]
polar_stat_stale_cost 0.0001 N [0-2147483647]
polar_auditlog_max_query_length 4096 N [512-49152]
polar_use_statistical_relpages on N [on|off]
shared_preload_libraries 'polar_vfs,polar_worker,dbms_pipe,polar_gen,pg_stat_statements,auth_delay,auto_explain,pg_cron,dbms_job,polar_stat_sql' Y ^'(polar_vfs,polar_worker)(,\w+)*'$
wal_keep_segments 32 N [0-100000]
temp_tablespaces 'polar_tmp' N
polar_csn_enable on Y [on|off]
polar_max_non_super_conns 400 N [-1-262143]
polar_enable_master_xlog_read_ahead on Y [on|off]
polar_enable_early_launch_checkpointer on Y [on|off]
polar_enable_early_launch_parallel_bgwriter on Y [on|off]
polar_enable_wal_prefetch off Y [on|off]
polar_enable_persisted_buffer_pool on Y [on|off]
polar_cast_decode_list '' N