Skip to main content
The Options type provides configuration parameters for fine-tuning the behavior of the Tashi Vertex consensus engine. It includes settings for timing, throughput, latency management, and network behavior.

​
Constructor

​
new

Initializes a default set of options. 
pub fn new() -> Self

​
Returns

​
return
Options
A new Options instance with default values.

​
Example

use tashi_vertex::Options;

let mut options = Options::new();
Options implements the Default trait, so you can also use Options::default().

​
Configuration methods

​
Timing and intervals

​
set_base_min_event_interval_us

Sets the base minimum event interval (in microseconds). 
pub fn set_base_min_event_interval_us(&mut self, interval: u64)
​
interval
u64
required
The minimum interval between events in microseconds

​
set_heartbeat_us

Sets the heartbeat interval (in microseconds). 
pub fn set_heartbeat_us(&mut self, heartbeat: u64)
​
heartbeat
u64
required
The heartbeat interval in microseconds. Default: 500,000 (500 milliseconds).
When there is no data to finalize, empty events are created at this interval to keep the session alive.

​
get_base_min_event_interval_us

Gets the base minimum event interval (in microseconds). 
pub fn get_base_min_event_interval_us(&self) -> u64

​
get_heartbeat_us

Gets the heartbeat interval (in microseconds). 
pub fn get_heartbeat_us(&self) -> u64

​
Latency management

​
set_target_ack_latency_ms

Sets the target acknowledgment latency (in milliseconds). 
pub fn set_target_ack_latency_ms(&mut self, latency: u32)
​
latency
u32
required
The target acknowledgment latency in milliseconds. Default: 400 milliseconds.
As throughput across the session increases, the ack latency increases. When the ack latency rises above this threshold, the node votes that throughput should not increase further.
If this threshold is lower than your uncongested ping, the node will erroneously always vote to restrict throughput.

​
set_max_ack_latency_ms

Sets the maximum acknowledgment latency (in milliseconds). 
pub fn set_max_ack_latency_ms(&mut self, latency: u32)
​
latency
u32
required
The maximum acknowledgment latency in milliseconds. Default: 600 milliseconds.
If ack latency rises above this threshold, the node votes to gradually reduce throughput across the session to bring it down.

​
set_throttle_ack_latency_ms

Sets the throttle acknowledgment latency (in milliseconds). 
pub fn set_throttle_ack_latency_ms(&mut self, latency: u32)
​
latency
u32
required
The throttle acknowledgment latency in milliseconds. Default: 900 milliseconds.
If ack latency rises above this threshold, the node votes to drastically restrict throughput across the session as an emergency measure.

​
set_reset_ack_latency_ms

Sets the reset acknowledgment latency (in milliseconds). 
pub fn set_reset_ack_latency_ms(&mut self, latency: u32)
​
latency
u32
required
The reset acknowledgment latency in milliseconds. Default: 2000 milliseconds.
If ack latency rises above this threshold, the node votes to reset throughput restriction to its initial value. This is a last-ditch effort to recover from rising ack latency.

​
get_target_ack_latency_ms

Gets the target acknowledgment latency (in milliseconds). 
pub fn get_target_ack_latency_ms(&self) -> u32

​
get_max_ack_latency_ms

Gets the maximum acknowledgment latency (in milliseconds). 
pub fn get_max_ack_latency_ms(&self) -> u32

​
get_throttle_ack_latency_ms

Gets the throttle acknowledgment latency (in milliseconds). 
pub fn get_throttle_ack_latency_ms(&self) -> u32

​
get_reset_ack_latency_ms

Gets the reset acknowledgment latency (in milliseconds). 
pub fn get_reset_ack_latency_ms(&self) -> u32

​
Peer management

​
set_fallen_behind_kick_s

Sets the number of seconds a creator can fall behind before being kicked. 
pub fn set_fallen_behind_kick_s(&mut self, kick: i64)
​
kick
i64
required
The number of seconds. If negative, the node will never vote to kick.
If a creator falls behind for this many seconds or more, the node will vote to kick them.

​
get_fallen_behind_kick_s

Gets the number of seconds a creator can fall behind before being kicked. 
pub fn get_fallen_behind_kick_s(&self) -> i64

​
Event reporting

​
set_report_gossip_events

Enables or disables reporting of gossip events. 
pub fn set_report_gossip_events(&mut self, enabled: bool)
​
enabled
bool
required
Whether to report gossip events

​
get_report_gossip_events

Gets whether reporting of gossip events is enabled. 
pub fn get_report_gossip_events(&self) -> bool

​
Dynamic epoch sizing

​
set_enable_dynamic_epoch_size

Enables or disables dynamic epoch sizing. 
pub fn set_enable_dynamic_epoch_size(&mut self, enabled: bool)
​
enabled
bool
required
Whether to enable dynamic epoch sizing. Default: true.
If true, the node will vote to resize the epoch depending on network conditions. Depending on network conditions, rounds may pass more quickly or more slowly. Whenever a creator joins or leaves, they’ll have to wait out the epoch before the address book change takes effect. A leaving creator doesn’t want to wait too long, and a joining creator needs a sufficiently long window in which to join. Creators who don’t disable this config option will automatically vote to keep epoch lengths in the range of 1 to 3 seconds.

​
get_enable_dynamic_epoch_size

Gets whether dynamic epoch sizing is enabled. 
pub fn get_enable_dynamic_epoch_size(&self) -> bool

​
Transaction management

​
set_transaction_channel_size

Sets the transaction channel size. 
pub fn set_transaction_channel_size(&mut self, size: usize)
​
size
usize
required
The maximum number of transactions to buffer. Default: 32.
The maximum number of transactions to buffer before applying backpressure.

​
set_max_unacknowledged_bytes

Sets the maximum number of unacknowledged bytes. 
pub fn set_max_unacknowledged_bytes(&mut self, bytes: usize)
​
bytes
usize
required
The maximum number of unacknowledged bytes. Default: 524,288,000 (500 MiB).
How many bytes worth of transactions that haven’t yet been seen by the network to pull from the transaction buffer.

​
get_transaction_channel_size

Gets the transaction channel size. 
pub fn get_transaction_channel_size(&self) -> usize

​
get_max_unacknowledged_bytes

Gets the maximum number of unacknowledged bytes. 
pub fn get_max_unacknowledged_bytes(&self) -> usize

​
Thread pool configuration

​
set_max_blocking_verify_threads

Sets the maximum number of blocking verify threads. 
pub fn set_max_blocking_verify_threads(&mut self, threads: usize)
​
threads
usize
required
The maximum number of threads. Default: number of CPU cores available.
Above a constant threshold, signature verifications are sent to a blocking thread pool instead of using spare compute time in Tokio’s core thread pool. This sets the maximum number of threads to spawn for blocking verifications.
This value cannot be zero or else events that grow larger than the threshold cannot be verified.

​
get_max_blocking_verify_threads

Gets the maximum number of blocking verify threads. 
pub fn get_max_blocking_verify_threads(&self) -> usize

​
State sharing

​
set_enable_state_sharing

Enables or disables state sharing. 
pub fn set_enable_state_sharing(&mut self, enabled: bool)
​
enabled
bool
required
Whether to enable state sharing. Default: false.

​
set_epoch_states_to_cache

Sets the number of epoch states to cache. 
pub fn set_epoch_states_to_cache(&mut self, epochs: u16)
​
epochs
u16
required
The number of epoch states to cache. Default: 3.
When state sharing is enabled, this is the number of epoch states to cache. If a fallen behind creator fails to download an epoch’s state in time, they will have to restart the download.

​
get_enable_state_sharing

Gets whether state sharing is enabled. 
pub fn get_enable_state_sharing(&self) -> bool

​
get_epoch_states_to_cache

Gets the number of epoch states to cache. 
pub fn get_epoch_states_to_cache(&self) -> u16

​
Network features

​
set_enable_hole_punching

Enables or disables hole punching. 
pub fn set_enable_hole_punching(&mut self, enabled: bool)
​
enabled
bool
required
Whether to enable hole punching. Default: true.
If true, the node will attempt to use UDP hole punching to establish direct connections between creators behind NATs.

​
get_enable_hole_punching

Gets whether hole punching is enabled. 
pub fn get_enable_hole_punching(&self) -> bool

​
Usage example

use tashi_vertex::Options;

let mut options = Options::new();

// Configure timing
options.set_heartbeat_us(500_000); // 500ms

// Configure latency thresholds
options.set_target_ack_latency_ms(400);
options.set_max_ack_latency_ms(600);
options.set_throttle_ack_latency_ms(900);

// Configure transaction handling
options.set_transaction_channel_size(64);
options.set_max_unacknowledged_bytes(1_000_000_000); // 1 GB

// Enable features
options.set_enable_dynamic_epoch_size(true);
options.set_enable_hole_punching(true);
options.set_enable_state_sharing(true);
options.set_epoch_states_to_cache(5);

// Query current values
let heartbeat = options.get_heartbeat_us();
let max_latency = options.get_max_ack_latency_ms();
All setter methods modify the options in place. Configure all desired options before passing to Engine::start().
  • Engine - Consumes the options when starting the engine

Build docs developers (and LLMs) love

Get started for freeTalk to us