ai_watchdog_monitor_status

metric_name

text

Metric names:

• tps: TPS.
• tps_hourly: average TPS per hour.
• shared_used_mem: the used shared memory (MB).
• dynamic_used_shrctx: the used shared memory context (MB).
• other_used_mem: other used memory (MB).
• process_used_mem: the used resident memory (MB).
• dynamic_used_mem: the used dynamic memory (MB).
• malloc_failures: the number of memory allocation failures in each collection interval.
• D_state_rate: percentage of threads in the D state.
• R_state_rate: percentage of threads in the R state.
• S_state_rate: ratio of threads in the S state.
• db_state: database state (68 indicates D, 82 indicates R, and 83 indicates S).
• cpu_usage: CPU usage. The upper limit is 100.
• disk_io: disk I/O delay between two collection intervals.
• network_io: network I/O delay between two collection intervals.
• threadpool_usage: thread pool usage.
• threadpool_hang_rate: percentage of the thread pool group in the hang state.

max_length

int

Collection queue length.

current_length

int

The number of currently collected samples.

collection_interval

int

Collection interval, in seconds.

latest_value

int

Value collected last time. If no value is collected, the value is null.

last_report

timestamp

The last collection time.