GetStatsByServer¶

Description:
Returns historical host metrics, the current host state and the latest internal platform runtime snapshot. Platform metrics describe the whole st-server process and are not divided by brand.

Access Control¶

🛡️ Access Level Required: SESSION_MANAGER, SESSION_ADMIN, SESSION_DEALER

Access Scope¶

The server resolves the manager from the current staff cache and requires admin scope or access_backoffice. Token payload is used only to resolve the manager id and session type.

Parameters¶

Name	Type	Required	Description
from	int	✅	Start timestamp (UNIX)
to	int	✅	End timestamp (UNIX)

Response¶

A response contains:

structure: ordered list of metric field names
rows: ordered metric snapshots for the requested interval plus one live snapshot row appended at the end
info: current server state summary
platform: latest internal platform snapshot collected by MetricsManager

The platform object is a live snapshot rather than a historical row. It is normally refreshed every 30 seconds. Consumers should inspect sampled_at and age_sec before treating its values as current.

`rows` / `structure` fields¶

Field	Type	Description
connections	int	Number of active connections counted during the sampling window
requests	int	Number of handled requests counted during the sampling window
cpu_load_pct	float	Total CPU usage in percent
disk_load_pct	float	Filesystem usage in percent for `/`
net_pck_in	float	Incoming packets per second on the monitored network interface
net_pck_out	float	Outgoing packets per second on the monitored network interface
ram_load_pct	float	RAM usage in percent
time	int	UNIX timestamp of the metric snapshot
cpu_irq_pct	float	CPU time spent in IRQ handlers in percent
cpu_softirq_pct	float	CPU time spent in soft IRQ handlers in percent
net_bytes_in	float	Incoming bytes per second on the monitored network interface
net_bytes_out	float	Outgoing bytes per second on the monitored network interface
disk_read_mbps	float	Disk read throughput in MB/s for the first detected block device
disk_write_mbps	float	Disk write throughput in MB/s for the first detected block device
sockets_total	int	Total observed TCP sockets
sockets_waiting	int	TCP sockets currently waiting in SYN states

`info` fields¶

Field	Type	Description
os	string	Operating system pretty name
cpu_name	string	CPU model name
cpu_cores	int	Number of CPU cores
cpu_load_pct	float	Current CPU usage in percent
ram_free_mb	int	Currently available RAM in MB
ram_total_mb	int	Total RAM in MB
ram_largest_mb	int	Largest free contiguous RAM block in MB
disk_free_mb	int	Free disk space in MB for `/`
disk_total_mb	int	Total disk space in MB for `/`
disk_read_mbps	float	Current disk read throughput in MB/s
disk_write_mbps	float	Current disk write throughput in MB/s
net_iface	string	Monitored network interface name
net_bytes_in	float	Current incoming bytes per second
net_bytes_out	float	Current outgoing bytes per second
net_pck_in	float	Current incoming packets per second
net_pck_out	float	Current outgoing packets per second
sockets_total	int	Current total TCP sockets
sockets_waiting	int	Current TCP sockets in waiting/SYN states

`platform` fields¶

Field	Type	Description
sampled_at	int	UNIX timestamp when the snapshot was collected
sequence	uint64	Monotonically increasing snapshot sequence within the current server process
age_sec	int	Snapshot age in seconds; `-1` means that the first snapshot is not available yet
runtime	object	Process memory, file descriptor and thread counters
quotes	object	Quote storage queue and sampled processing latency
event_bus	object	Local EventBus state
thread_pools	object	Tracked task state for the general and quote executors
trade_pipeline	object	Internal trade calculation and persistence queues
feeders	array	Runtime state of configured quote feeders

`platform.runtime`¶

Field	Type	Description
rss_kb	uint64	Resident memory in KiB
rss_anon_kb	uint64	Anonymous resident memory in KiB
vmdata_kb	uint64	Process data virtual memory in KiB
heap_arena_bytes	uint64	Bytes currently allocated to heap arenas
heap_used_bytes	uint64	Used heap bytes reported by the allocator
heap_free_bytes	uint64	Free bytes retained inside heap arenas
heap_mmap_bytes	uint64	Heap bytes backed by separate mmap allocations
heap_mmap_blocks	uint64	Number of mmap-backed heap allocations
fd_total	uint64	Total open file descriptors
fd_sockets	uint64	File descriptors referring to sockets
fd_files	uint64	File descriptors referring to regular files
threads	uint64	Number of process threads

`platform.quotes`¶

Field	Type	Description
ticks_enqueued	uint64	Quote storage jobs accepted since process start
ticks_dequeued	uint64	Quote storage jobs processed since process start
ticks_dropped	uint64	Quote storage jobs rejected since process start
backlog	uint64	Current quote storage queue backlog
max_worker_backlog	uint64	Highest observed worker backlog
latency_samples	uint64	Number of quote callbacks sampled during the latest snapshot interval
latency_p50_us	uint64	Approximate p50 processing latency for the latest interval, in microseconds
latency_p95_us	uint64	Approximate p95 processing latency for the latest interval, in microseconds
latency_p99_us	uint64	Approximate p99 processing latency for the latest interval, in microseconds
latency_max_us	uint64	Maximum sampled processing latency for the latest interval

Quote latency uses deterministic per-thread sampling: one callback out of 128 is measured. Unsampled callbacks do not call the clock and therefore avoid adding timing overhead to every quote. Percentiles are histogram approximations for the latest collection interval, not values split by symbol or brand.

`platform.event_bus`¶

Field	Type	Description
backlog	uint64	Approximate number of events awaiting dispatch
event_names	uint64	Number of registered event names
handlers	uint64	Number of registered handlers
running	bool	Whether the local EventBus worker is running

`platform.thread_pools.general` and `platform.thread_pools.quote`¶

Field	Type	Description
state	int	Pool lifecycle: `0` stopped, `1` running, `2` draining, `3` stopping
threads	uint64	Configured worker count
queued	uint64	Tracked tasks accepted but not started
running	uint64	Tracked tasks currently executing
completed	uint64	Tracked tasks completed since process start
rejected	uint64	Tasks rejected because the pool was unavailable or shutting down
failed	uint64	Tasks that raised an exception inside the tracked wrapper
slow	uint64	Tasks whose execution took at least five seconds
max_duration_ms	uint64	Maximum completed tracked-task duration
active_tasks	array	Currently executing tracked tasks with `id`, `name`, `thread_id`, enqueue/start timestamps, and `running_ms`

Task metadata never contains request bodies, credentials, or business payloads. It identifies only the internal operation, for example workflow.trade.closed, eventbus.account.online, plugin.event.1, or core.quote.publish.

`platform.trade_pipeline`¶

Field	Type	Description
batches_enqueued	uint64	Calculation batches queued since process start
batches_completed	uint64	Calculation batches completed since process start
batches_in_flight	uint64	Calculation batches currently running
trades_in_flight	uint64	Trade records currently being calculated
trades_high_watermark	uint64	Highest observed number of trades in flight
runtime_tasks_in_flight	uint64	Trade runtime tasks currently running
db_ops_queued	uint64	Trade persistence operations waiting in the database queue
db_tasks_in_flight	uint64	Trade persistence tasks currently running

`platform.feeders[]`¶

Field	Type	Description
index	int	Feeder index
name	string	Configured feeder name
state	int	Current feeder runtime state
ticks_total	uint64	Ticks received from the feeder since process start
tick_errors	uint64	Feeder callback errors since process start
callbacks_in_flight	uint64	Callbacks currently running
callbacks_high_watermark	uint64	Highest observed callbacks-in-flight value

Example Response¶

{
  "structure": [
    "connections",
    "requests",
    "cpu_load_pct",
    "disk_load_pct",
    "net_pck_in",
    "net_pck_out",
    "ram_load_pct",
    "time",
    "cpu_irq_pct",
    "cpu_softirq_pct",
    "net_bytes_in",
    "net_bytes_out",
    "disk_read_mbps",
    "disk_write_mbps",
    "sockets_total",
    "sockets_waiting"
  ],
  "rows": [
    [123, 4567, 23.5, 14.8, 120.0, 115.0, 67.1, 1712345678, 0.3, 0.8, 1048576.0, 943718.0, 12.4, 3.1, 842, 4]
  ],
  "info": {
    "os": "Ubuntu 24.04 LTS",
    "cpu_name": "Intel(R) Xeon(R) CPU",
    "cpu_cores": 8,
    "cpu_load_pct": 21.2,
    "ram_free_mb": 6144,
    "ram_total_mb": 16384,
    "ram_largest_mb": 512,
    "disk_free_mb": 102400,
    "disk_total_mb": 204800,
    "disk_read_mbps": 12.4,
    "disk_write_mbps": 3.1,
    "net_iface": "eth0",
    "net_bytes_in": 1048576.0,
    "net_bytes_out": 943718.0,
    "net_pck_in": 120.0,
    "net_pck_out": 115.0,
    "sockets_total": 842,
    "sockets_waiting": 4
  },
  "platform": {
    "sampled_at": 1789545722,
    "sequence": 125,
    "age_sec": 7,
    "runtime": {
      "rss_kb": 7355392,
      "rss_anon_kb": 7323648,
      "vmdata_kb": 23532544,
      "heap_arena_bytes": 21103640576,
      "heap_used_bytes": 20796407808,
      "heap_free_bytes": 307232768,
      "heap_mmap_bytes": 2232418304,
      "heap_mmap_blocks": 229,
      "fd_total": 466,
      "fd_sockets": 19,
      "fd_files": 429,
      "threads": 87
    },
    "quotes": {
      "ticks_enqueued": 169238,
      "ticks_dequeued": 169238,
      "ticks_dropped": 0,
      "backlog": 0,
      "max_worker_backlog": 0,
      "latency_samples": 55,
      "latency_p50_us": 500,
      "latency_p95_us": 2000,
      "latency_p99_us": 5000,
      "latency_max_us": 4312
    },
    "event_bus": {
      "backlog": 0,
      "event_names": 22,
      "handlers": 31,
      "running": true
    },
    "thread_pools": {
      "general": {
        "state": 1,
        "threads": 4,
        "queued": 0,
        "running": 1,
        "completed": 18432,
        "rejected": 0,
        "failed": 0,
        "slow": 1,
        "max_duration_ms": 6120,
        "active_tasks": [
          {
            "id": 18433,
            "name": "workflow.trade.closed",
            "thread_id": "140172902938368",
            "queued_at_ms": 1789545722000,
            "started_at_ms": 1789545722001,
            "running_ms": 37
          }
        ]
      },
      "quote": {
        "state": 1,
        "threads": 2,
        "queued": 0,
        "running": 0,
        "completed": 95431,
        "rejected": 0,
        "failed": 0,
        "slow": 0,
        "max_duration_ms": 14,
        "active_tasks": []
      }
    },
    "trade_pipeline": {
      "batches_enqueued": 40006,
      "batches_completed": 40006,
      "batches_in_flight": 0,
      "trades_in_flight": 0,
      "trades_high_watermark": 547946,
      "runtime_tasks_in_flight": 0,
      "db_ops_queued": 0,
      "db_tasks_in_flight": 0
    },
    "feeders": [
      {
        "index": 7,
        "name": "SCT-Feeder",
        "state": 1,
        "ticks_total": 159293,
        "tick_errors": 0,
        "callbacks_in_flight": 0,
        "callbacks_high_watermark": 1
      }
    ]
  }
}

Collection and provider monitoring¶

The platform snapshot has three consumers:

MetricsManager exposes the current snapshot through this BO method.
With server debug logging enabled, the same runtime values are printed as [MONITORING][...] debug messages.
The server exposes Prometheus text through the local Unix socket /run/sttrader/metrics.sock. stmonit reads this socket, adds instance-level labels and sends the metrics to the provider's centralized registry.

The Unix socket does not open an inbound TCP port. Failure or absence of stmonit does not block quote processing, trade calculation or BO access.

Errors¶

Code	Description
400	Invalid request data
500	Internal server error

GetStatsByServer¶

Access Control¶

Access Scope¶

Parameters¶

Response¶

rows / structure fields¶

info fields¶

platform fields¶

platform.runtime¶

platform.quotes¶

platform.event_bus¶

platform.thread_pools.general and platform.thread_pools.quote¶

platform.trade_pipeline¶

platform.feeders[]¶