beryl/stats

Local coordinator statistics.

Snapshots are point-in-time values captured when the local coordinator services a request. They are intended for operational polling, not as an event stream. Poll no more frequently than roughly once per second so observation does not add meaningful coordinator load.

Types

Snapshot

A point-in-time snapshot of local coordinator state.

pub type Snapshot

SnapshotError

Errors returned while requesting a coordinator snapshot.

pub type SnapshotError {
  CoordinatorUnavailable
  RequestTimedOut
}

Constructors

CoordinatorUnavailable

The local coordinator is not currently running.

RequestTimedOut

The coordinator did not service the request within the bounded timeout.

Functions

active_topics

Return the number of topics with at least one local joined socket.

pub fn active_topics(Snapshot) -> Int

connected_sockets

Return the number of sockets connected to the local coordinator.

pub fn connected_sockets(Snapshot) -> Int

coordinator_mailbox_length

Return the coordinator mailbox length when it serviced the request.

pub fn coordinator_mailbox_length(Snapshot) -> Int

joined_socket_topic_pairs

Return the number of joined socket/topic pairs.

One socket joined to two topics contributes two pairs.

pub fn joined_socket_topic_pairs(Snapshot) -> Int

registered_channel_handlers

Return the number of channel handlers registered with the coordinator.

pub fn registered_channel_handlers(Snapshot) -> Int

snapshot

Request a point-in-time snapshot from the local coordinator.

The request waits for at most approximately one second. During a coordinator restart this returns CoordinatorUnavailable or RequestTimedOut; an overloaded coordinator returns RequestTimedOut. Neither condition panics. This API reports only the node represented by channels; aggregate multi-node statistics outside Beryl.

Poll no more frequently than roughly once per second.

pub fn snapshot(beryl.Channels) -> Result(Snapshot, SnapshotError)