Help Center/ GaussDB/ Distributed_8.x/ Schemas/ DBE_PERF Schema/ Utility/ GLOBAL_REDO_STATUS

Updated on 2024-06-03 GMT+08:00

View PDF

GLOBAL_REDO_STATUS

GLOBAL_REDO_STATUS displays the replaying of logs about instances in the entire cluster, as described in Table 1.

**Table 1** GLOBAL_REDO_STATUS columns
Name	Type	Description
node_name	text	Instance name.
redo_start_ptr	bigint	Start point for replaying the instance logs.
redo_start_time	bigint	Start time (UTC) when the instance logs are replayed.
redo_done_time	bigint	End time (UTC) when the instance logs are replayed.
curr_time	bigint	Current time (UTC) of the instance.
min_recovery_point	bigint	Position of the minimum consistency point for the instance logs.
read_ptr	bigint	Position for reading the instance logs.
last_replayed_read_ptr	bigint	Position for replaying the instance logs.
recovery_done_ptr	bigint	Replay position after the instance is started.
read_xlog_io_counter	bigint	Number of I/Os when the instance reads and replays logs.
read_xlog_io_total_dur	bigint	Total I/O latency when the instance reads and replays logs.
read_data_io_counter	bigint	Number of data page I/O reads during replay in the instance.
read_data_io_total_dur	bigint	Total I/O latency of data page reads during replay in the instance.
write_data_io_counter	bigint	Number of data page I/O writes during replay in the instance.
write_data_io_total_dur	bigint	Total I/O latency of data page writes during replay in the instance.
process_pending_counter	bigint	Number of synchronization times of log distribution threads during replay in the instance.
process_pending_total_dur	bigint	Total synchronization latency of log distribution threads during replay in the instance.
apply_counter	bigint	Number of synchronization times of replay threads during replay in the instance.
apply_total_dur	bigint	Total synchronization latency of replay threads during replay in the instance.
speed	bigint	Log replay rate of the current instance. The value is updated every time when a 256 MB log is replayed. The unit is byte/s. In a cluster environment, you are advised to run the cm_ctl query -rv command to obtain a more accurate replay speed of the standby node. For details about the cm_ctl command, see "Tools Used in the System > cm_ctl" in the Tool Reference.
local_max_ptr	bigint	Maximum number of replay logs received by the local host after the instance is started.
primary_flush_ptr	bigint	Log point where the host flushes logs to a disk.
worker_info	text	Replay thread information of the instance. If concurrent replay is not enabled, the value is NULL.