metrics

Hyperliquid Exporter Metrics

Here's some information about each metric exposed by the Hyperliquid Exporter, along with sample Prometheus queries for analysis.

Block Metrics

hl_block_height

• Type: Gauge
• Description: The current block height of the Hyperliquid blockchain.
• Usage: This metric helps track the progression of the blockchain and can be used to monitor chain liveness.

Sample queries

#Block height increase rate (blocks per second) over the last hour

rate(hl_block_height[1h])

hl_latest_block_time

• Type: Gauge
• Description: The Unix timestamp of the latest block.
• Usage: This metric is useful for monitoring chain liveness and calculating time-based statistics.

Sample queries:

# latest block time (use unit datetime)
hl_latest_block_time * 1000

hl_apply_duration_milliseconds

• Type: Histogram
• Description: Distribution of block apply durations in milliseconds.

Buckets:

[0.1, 0.2, 0.5, 1, 2, 3, 5, 7, 10, 15, 20,
30, 50, 75, 100, 150, 160, 170, 180, 190, 
200, 210, 220, 230, 240, 250]

Usage: Provides insights into the performance of block application which can help identify potential bottlenecks or performance issues.

Sample queries

# Median apply duration over the last 5 minutes
histogram_quantile(0.5, rate(hl_apply_duration_milliseconds_bucket[5m]))

# 90th percentile apply duration over the last 5 minutes
histogram_quantile(0.9, rate(hl_apply_duration_milliseconds_bucket[5m]))

# Percentage of apply durations under 10ms
sum(rate(hl_apply_duration_milliseconds_bucket{le="10"}[5m])) / sum(rate(hl_apply_duration_milliseconds_count[5m])) * 100

# Average apply duration
rate(hl_apply_duration_milliseconds_sum[5m]) / rate(hl_apply_duration_milliseconds_count[5m])

# Distribution of apply durations
rate(hl_apply_duration_milliseconds_bucket[5m])

# Count of apply durations exceeding 1 second
sum(rate(hl_apply_duration_milliseconds_bucket{le="+Inf"}[5m])) - sum(rate(hl_apply_duration_milliseconds_bucket{le="1"}[5m]))

hl_block_time_milliseconds

Type: Histogram Description: Histogram of time between blocks in milliseconds.

Buckets:

[10, 20, 30, 40, 50, 60, 70, 80, 90, 100,
120, 140, 160, 180, 200, 220, 240, 260, 280, 300,
300, 350, 400, 450, 500, 600, 700, 800,
900, 1000, 1500, 2000]

Usage: Similar queries as to hl_apply_duration_milliseconds can be applied

hl_proposer_count_total

• hl_proposer_count_total
• Type: Counter
• Description: The total number of blocks proposed by each validator address.
• Labels: proposer (validator address)

Usage: This metric helps monitor the distribution of block proposals among validators and can be used to detect potential issues with specific validators.

Sample Queries:

# Top 5 proposers by number of proposed blocks
topk(5, hl_proposer_count_total)

# Percentage of blocks proposed by each validator
hl_proposer_count_total / sum(hl_proposer_count_total) * 100

# Proposal rate (blocks per minute) for each validator over the last hour
rate(hl_proposer_count_total[1h]) 60

# Alert if a validator hasn't proposed a block in the last 24 hours
changes(hl_proposer_count_total[24h]) == 0

hl_software_version_info

• Type: Gauge
• Description: Information about the current software version running on the node.

Labels:

• commit: Git commit hash of the current version
• date: Build date of the current version

Usage: This metric helps track software versions across nodes and can be used to ensure all nodes are running the expected version.

Sample Queries:

# Check if all nodes are running the same version
count(count by (commit) (hl_software_version_info)) > 1

# List all unique versions currently running
count by (commit, date) (hl_software_version_info)

Stake Status

hl_total_stake

• Type: Gauge
• Description: Total stake across all validators in the Hyperliquid network.
• Usage: This metric provides an overview of the total stake in the network, which can be used to assess the overall security and economic value locked in the network.

Sample Queries

# Current total stake
hl_total_stake

# Percentage change in total stake over the last 24 hours
(hl_total_stake - hl_total_stake offset 24h) / hl_total_stake offset 24h * 100

hl_jailed_stake

• Type: Gauge
• Description: Total stake of jailed validators in the Hyperliquid network.
• Usage: This metric helps monitor the amount of stake that is currently jailed, which can indicate network health and validator behavior.

Sample Queries

# Percentage of total stake that is jailed
hl_jailed_stake / hl_total_stake * 100

# Alert if jailed stake exceeds 10% of total stake
hl_jailed_stake / hl_total_stake > 0.1

hl_not_jailed_stake

• Type: Gauge
• Description: Total stake of not jailed validators in the Hyperliquid network.
• Usage: This metric represents the active stake in the network, which is crucial for understanding the current voting power distribution. Sample Queries:

# Ratio of not jailed stake to total stake
hl_not_jailed_stake / hl_total_stake

# Trend of not jailed stake over the last week
rate(hl_not_jailed_stake[1w])

hl_validator_stake

• Type: Gauge
• Description: Individual stake of each validator in the Hyperliquid network.
• Labels: validator (validator address)
• Usage: This metric allows for tracking the stake of individual validators, which is useful for monitoring validator performance and network decentralization.

Sample Queries

# Top 10 validators by stake
topk(10, hl_validator_stake)

# Validators with less than 1% of total stake
hl_validator_stake / ignoring(validator) group_left() hl_total_stake < 0.01

# Stake concentration: percentage of stake held by top 5 validators
sum(topk(5, hl_validator_stake)) / hl_total_stake * 100

hl_validator_jailed_status

• Type: Gauge
• Description: Jailed status of each validator (1 if jailed, 0 if not jailed).
• Labels: validator (validator address)
• Usage: This metric helps monitor the jailed status of validators, which is crucial for understanding network health and validator behavior.

Sample Queries

# Count of jailed validators
sum(hl_validator_jailed_status == 1)

# List of jailed validators
hl_validator_jailed_status == 1

# Alert if a top 10 validator by stake is jailed
hl_validator_jailed_status{validator=~"$top_10_validators"} == 1

hl_validator_count

• Type: Gauge
• Description: Total number of validators in the Hyperliquid network.
• Usage: This metric provides insight into the size and decentralization of the validator set.

Sample Queries

# Current validator count
hl_validator_count

# Change in validator count over the last 24 hours
hl_validator_count - hl_validator_count offset 24h

# Alert if validator count drops below a threshold
hl_validator_count < 50