Use native backend proof for mining status

.env.example

@@ -33,7 +33,7 @@ BDAG_NODE_ARCHIVAL=0
 # Dashboard repository. Release builds always use the main branch.
 DASHBOARD_REPO=https://github.com/BlockdagEngineering/redis-dash.git
 
-P2P_PORT=8150
+P2P_PORT=8155
 BDAG_RPC_PORT=38131
 EVM_HTTP_PORT=18545
 EVM_WS_PORT=18546
@@ -45,12 +45,14 @@ DASHBOARD_HOST_PORT=8088
 # Dashboard UI display timezone.
 DASHBOARD_TIMEZONE=Africa/Johannesburg
 
-# NODE_DATA_DIR=./node-data
+NODE_DATA_DIR=./data/node
 
-# Required for any pool/miner deployment. Do not use the zero address.
+# Required for any pool/miner deployment. Do not use the zero address. Leave
+# blank in the example so release/install readiness fails until an operator
+# supplies the payout address for this deployment.
 # POOL_COINBASE_ADDRESS may be set separately when it must differ from the
 # displayed miner/pool identity; otherwise set MINING_POOL_ADDRESS only.
-MINING_POOL_ADDRESS=0xEd5e2dD1377aBe28907673cdEa8c2c04Fb715dcd
+MINING_POOL_ADDRESS=
 POOL_COINBASE_ADDRESS=
 
 # Storage placement for constrained mining appliances. Release installers keep
@@ -87,9 +89,12 @@ BDAG_P2P_PROTOCOLS=tcp
 # Standard mining-host performance defaults. The release topology has one
 # production node with a direct pool-to-node RPC path.
 BDAG_NODE_CACHE_MB=2048
+BDAG_NODE_DEBUG_LEVEL=warn
+BDAG_NODE_NO_FILE_LOGGING=1
 BDAG_EVM_CACHE_MB=2048
 BDAG_NODE_CACHE_DATABASE_PERCENT=70
 BDAG_NODE_OBSOLETE_HEIGHT=20
+BDAG_NODE_MINING_NO_PENDING_TX=0
 BDAG_BLOCK_READ_AHEAD_KB=1024
 BDAG_BLOCK_NR_REQUESTS=256
 BDAG_NODE_CPU_SHARES=6144
@@ -127,13 +132,18 @@ POOL_GBT_PRESSURE_INTERVAL_MS=500
 POOL_GBT_PRESSURE_WINDOW_SECONDS=10
 POOL_RPC_ROUTER_NODE_HEALTH_PROBE_SECONDS=15
 POOL_RPC_ROUTER_NODE_HEALTH_MAX_AGE_SECONDS=30
+POOL_RPC_ROUTER_NODE_HEALTH_FRESH_TEMPLATE_GRACE_SECONDS=15
+POOL_RPC_ROUTER_NODE_HEALTH_MIN_CONSENSUS_PEERS=2
 
 # Host network discovery is only for choosing this node's advertised P2P address.
 # Peer candidates should be complete P2P multiaddrs.
 BDAG_NETWORK_TOPOLOGY=auto
 BDAG_DETECTED_NETWORK_TOPOLOGY=
 BDAG_ASIC_LAN_INTERFACE=
 BDAG_ASIC_LAN_CIDRS=
+POOL_ASIC_ARP_TABLE_PATH=/host/proc/net/arp
+POOL_ASIC_MAC_ALLOWLIST=
+POOL_ASIC_MAC_OVERRIDES=
 # Optional route to miners on a remote LAN reached through a routed gateway
 # (e.g. a ZeroTier peer). When set, the miner-route compose service keeps the
 # host route alive so the watchdog can reach the miner HTTP API. Leave empty
@@ -151,9 +161,14 @@ BDAG_CHAIN_PEERSTORE_PEER_EXTRACTION_ENABLED=1
 BDAG_CHAIN_PEERSTORE_LOG_TAIL=8000
 EXTRA_PEER_ADDRESSES=
 
-# When enabled by installer/watchdog for attached ASICs, this must include
-# --miner, --miningaddr=<non-zero wallet>, and --maxinbound=1 on constrained
-# USB/router hosts. Do not add --allowminingwhennearlysynced or
+# Node template generation is enabled for this mining stack. When
+# BDAG_NODE_MINING_ARGS is blank, the node entrypoint derives
+# --miner --miningaddr=<POOL_COINBASE_ADDRESS|MINING_POOL_ADDRESS> at container
+# start. Catch-up repairs may explicitly set BDAG_ENABLE_NODE_MINING=0.
+BDAG_ENABLE_NODE_MINING=0
+BDAG_NODE_MODULES=Blockdag,miner
+# When set explicitly, this must include --miner and
+# --miningaddr=<non-zero wallet>. Do not add --allowminingwhennearlysynced or
 # --allowsubmitwhennotsynced; template/readiness health must fail closed when
 # P2P freshness or sync safety is unavailable.
 BDAG_NODE_MINING_ARGS=
@@ -175,6 +190,7 @@ BDAG_ADAPTIVE_CHAIN_RPC_WARN_MS=1000
 BDAG_STATUS_SAMPLER_ENABLED=1
 BDAG_STATUS_SAMPLER_INTERVAL_SECONDS=10
 BDAG_STATUS_SAMPLER_MAX_AGE_SECONDS=120
+BDAG_WATCHDOG_NATIVE_P2P_PEER_LOSS_REPAIR_SECONDS=300
 BDAG_CHAIN_STATE_SELF_HEAL_ENABLED=1
 BDAG_CHAIN_STATE_SELF_HEAL_UNIT=bdag-chain-state-self-heal.service
 BDAG_MINING_IMPERATIVE_CHAIN_STATE_RESTORE_ENABLED=1
@@ -196,11 +212,12 @@ BDAG_WATCHDOG_EARNINGS_SNAPSHOT_INTERVAL_SECONDS=60
 BDAG_STATUS_SAMPLER_EARNINGS_SNAPSHOT_INTERVAL_SECONDS=60
 BDAG_DASHBOARD_EARNINGS_SAMPLER_INTERVAL_SECONDS=60
 BDAG_DASHBOARD_GLOBAL_SAMPLER_INTERVAL_SECONDS=60
+BDAG_EVM_SYNC_BACKOFF_SECONDS=60
 BDAG_OPS_UID=1000
 BDAG_OPS_GID=1000
 BDAG_DOCKER_SOCKET_GID=0
 BDAG_GLOBAL_CACHE_TTL_SECONDS=60
-BDAG_GLOBAL_RPC_WORKERS=24
+BDAG_GLOBAL_RPC_WORKERS=4
 BDAG_GLOBAL_BLOCK_WINDOW=600
 BDAG_GLOBAL_CACHE_MAX_TIP_LAG_BLOCKS=30
 BDAG_INSTALL_REBUILD_DASHBOARD_PLOTS=1
@@ -424,18 +441,30 @@ MINER_WORKERS=1
 
 
 
-# Conservative block-candidate policy for constrained nodes. Keep template
-# refresh/recovery slower than the node RPC client ceiling can absorb.
+# ASIC block-candidate timing pinned to the best observed production window.
 POOL_RPC_ROUTER_NODE_HEALTH_ENABLED=true
-POOL_TEMPLATE_TTL_REFRESH_MS=2000
+POOL_RPC_ROUTER_EVM_HEAD_GUARD_ENABLED=false
+POOL_TEMPLATE_TTL_REFRESH_MS=100
 POOL_SUBMIT_STALE_BLOCK_CANDIDATES=false
-POOL_MAX_BLOCK_CANDIDATE_JOB_AGE_MS=2500
+POOL_RECENT_STALE_BLOCK_CANDIDATE_SUBMIT_GRACE_MS=1750
+POOL_MAX_BLOCK_CANDIDATE_JOB_AGE_MS=1750
+POOL_BLOCK_TIMING_CONTROLLER_MIN_JOB_AGE_MS=1750
+POOL_BLOCK_TIMING_CONTROLLER_MAX_JOB_AGE_MS=1750
+POOL_BLOCK_TIMING_CONTROLLER_MIN_TEMPLATE_TTL_MS=100
+POOL_BLOCK_TIMING_CONTROLLER_MAX_TEMPLATE_TTL_MS=100
+POOL_AUTO_TUNE_BLOCK_CANDIDATE_JOB_AGE=true
+POOL_AUTO_TUNE_BLOCK_CANDIDATE_MAX_AGE_MS=8000
 POOL_SUBMIT_BLOCK_HEADER_V2_ENABLED=true
 POOL_STALE_RACE_REJECT_WINDOW_SECONDS=10
 POOL_STALE_RACE_CLIENT_RESEND_THRESHOLD=1
+POOL_STALE_RACE_CLIENT_RECONNECT_THRESHOLD=0
 POOL_STALE_RACE_RECOVERY_COOLDOWN_SECONDS=5
 POOL_EXPIRED_JOB_CLIENT_RECONNECT_THRESHOLD=3
 POOL_EXPIRED_JOB_CLIENT_RECONNECT_WINDOW_SECONDS=120
 POOL_EXPIRED_JOB_CLIENT_RECONNECT_COOLDOWN_SECONDS=60
 POOL_ALLOW_MULTIPLE_BLOCK_CANDIDATES_PER_JOB=true
 POOL_PREEMPTIVE_BLOCK_CANDIDATE_REFRESH_ENABLED=true
+POOL_PREEMPTIVE_BLOCK_CANDIDATE_CLEAN_REISSUE_ENABLED=false
+POOL_PREEMPTIVE_BLOCK_CANDIDATE_REFRESH_DELAY_MS=0
+POOL_PREEMPTIVE_BLOCK_CANDIDATE_REFRESH_INTERVAL_MS=20
+POOL_PREEMPTIVE_BLOCK_CANDIDATE_REFRESH_TIMEOUT_MS=1000

README.md

@@ -291,12 +291,41 @@ is behind peers or while the selected backend is not mineable/submit-ready.
 using `iowait`, `io_some`, and `io_full` pressure signals; a production node
 more than `BDAG_CATCHUP_PAUSE_THRESHOLD_BLOCKS=300` blocks behind peers is the
 backup trigger when pressure signals are missing or delayed.
-The status sampler stops the pool, disables node mining/template runtime churn,
-raises the node cache toward `BDAG_CATCHUP_NODE_CACHE_MB` within the host memory
-budget, and recreates only the node service when that runtime change is needed.
-The dashboard reports this as a deliberate catch-up pause, not a pool failure,
-and tells operators to leave miners configured until I/O pressure drops, peer lag
-is back inside the safe window, and template health is ready.
+The status sampler leaves the pool container running for share/session
+continuity, pauses pool-side template work, disables node mining/template runtime
+churn, and may raise the node cache toward `BDAG_CATCHUP_NODE_CACHE_MB` within
+the host memory budget. It must not recreate the node while the pool is live or
+accepted-block history exists. The dashboard reports this as a deliberate
+catch-up pause, not a pool failure, and tells operators to leave miners
+configured until I/O pressure drops, peer lag is back inside the safe window, and
+template health is ready.
+
+Native P2P freshness remains the mining hard gate. A zero-peer native P2P sample
+blocks new mining work immediately, but `BDAG_WATCHDOG_NATIVE_P2P_PEER_LOSS_REPAIR_SECONDS=300`
+prevents the watchdog from restarting the node until peer loss is sustained and
+fresh paid-block evidence has expired. Chain-state restore is stricter than
+catch-up display state: it requires native peer-lag evidence and ignores
+EVM/public-reference lag by itself.
+
+EVM/public-reference lag is advisory only after native mining proof is complete:
+`getTemplateHealth` or backend metrics must prove current chain, fresh P2P,
+fresh consensus peer floor, peer lead inside the configured safety window,
+template readiness, and submit readiness. Unknown peer count or unknown peer
+lead fails closed. When miners are connected, zero ready miner lanes keeps
+`can_submit_blocks=false` unless fresh paid-block evidence proves the submit
+path is already converting work.
+
+Node config edits and service recreates are cold/idle operations. Before the
+pool is live, automation may prepare missing node mining/template support only
+after native safety gates pass. Once `pool` is running, or once any
+accepted-block history exists, automation must not rewrite node
+mining/template flags, peer lists, cache settings, or recreate `node`. Recovery
+should fail closed by pausing templates, preserving the existing node
+identity/data path, and waiting for native `getTemplateHealth` to prove
+`chain_current`, `p2p_mining_fresh`, `mineable_now`, and `submit_ready`. If
+chain data must be restored, stop the mining path deliberately, quarantine the
+old datadir, preserve `network.key`, restore a verified raw datadir or
+snapshot, and restart the existing `node` container without a Compose recreate.
 
 Dashboard block height is sourced from chain RPC `getBlockCount`; template
 height, logs, and main-order values are shown only as

docker-compose.yml

@@ -99,10 +99,20 @@ services:
       TMP: /tmp
       TEMP: /tmp
       BDAG_NETWORK: ${BDAG_NETWORK:-}
-      P2P_PORT: ${P2P_PORT:-8150}
+      P2P_PORT: ${P2P_PORT:-8155}
       BDAG_EPHEMERAL_DIR: /run/bdag-ephemeral
       BDAG_SNAPSHOT_URL: ${BDAG_SNAPSHOT_URL:-}
       BDAG_NODE_ARCHIVAL: ${BDAG_NODE_ARCHIVAL:-0}
+      BDAG_NODE_DEBUG_LEVEL: ${BDAG_NODE_DEBUG_LEVEL:-warn}
+      BDAG_NODE_NO_FILE_LOGGING: ${BDAG_NODE_NO_FILE_LOGGING:-1}
+      BDAG_NODE_OBSOLETE_HEIGHT: ${BDAG_NODE_OBSOLETE_HEIGHT:-20}
+      BDAG_NODE_MINING_NO_PENDING_TX: ${BDAG_NODE_MINING_NO_PENDING_TX:-0}
+      BDAG_ENABLE_NODE_MINING: ${BDAG_ENABLE_NODE_MINING:-0}
+      BDAG_NODE_MODULES: ${BDAG_NODE_MODULES:-Blockdag,miner}
+      BDAG_NODE_MINING_ARGS: ${BDAG_NODE_MINING_ARGS:-}
+      MINING_POOL_ADDRESS: ${MINING_POOL_ADDRESS:?set MINING_POOL_ADDRESS to a non-zero payout address}
+      POOL_COINBASE_ADDRESS: ${POOL_COINBASE_ADDRESS:-}
+      MINING_ADDRESS: ${MINING_ADDRESS:-}
       NODE_ARGS_APPEND: ${NODE_ARGS_APPEND:-}
       SYNC_SOURCE_NODE: ${SYNC_SOURCE_NODE:-0}
       BDAG_STORAGE_PROFILE: ${BDAG_STORAGE_PROFILE:-auto}
@@ -121,7 +131,7 @@ services:
     # so if you want to start fresh, you need to delete the volume explicity.
     # `docker compose down && docker volume rm <project>_node-data`\
 
-      - ${NODE_DATA_DIR:-node-data}:/var/lib/bdagStack/node
+      - ${NODE_DATA_DIR:-./data/node}:/var/lib/bdagStack/node
       - nodeworker-data:/var/lib/bdagStack/nodeworker
       - ${SNAPSHOT_HOST_PATH:-./docker/no-snapshot.marker}:/snapshot/latest.bdsnap:ro
       # node.conf in project root (copy from node.conf.example). gitignored.
@@ -169,24 +179,26 @@ services:
       POOL_FEE_PERCENTAGE: ${POOL_FEE_PERCENTAGE:-1.0}
       POOL_BLOCK_MATURITY: ${POOL_BLOCK_MATURITY:-10}
       POOL_PAYOUT_MATURITY: ${POOL_PAYOUT_MATURITY:-4096}
+      POOL_PAYOUT_MAX_BLOCKS_PER_TICK: ${POOL_PAYOUT_MAX_BLOCKS_PER_TICK:-5}
+      POOL_STRATUM_CONNECTION_LOG_INTERVAL_SECONDS: ${POOL_STRATUM_CONNECTION_LOG_INTERVAL_SECONDS:-60}
       POOL_RPC_ROUTER_NODE_HEALTH_ENABLED: ${POOL_RPC_ROUTER_NODE_HEALTH_ENABLED:-true}
       POOL_RPC_ROUTER_NODE_HEALTH_PROBE_SECONDS: ${POOL_RPC_ROUTER_NODE_HEALTH_PROBE_SECONDS:-15}
       POOL_RPC_ROUTER_NODE_HEALTH_MAX_AGE_SECONDS: ${POOL_RPC_ROUTER_NODE_HEALTH_MAX_AGE_SECONDS:-30}
       POOL_RPC_ROUTER_NODE_HEALTH_FRESH_TEMPLATE_GRACE_SECONDS: ${POOL_RPC_ROUTER_NODE_HEALTH_FRESH_TEMPLATE_GRACE_SECONDS:-15}
       POOL_RPC_ROUTER_NODE_HEALTH_MIN_CONSENSUS_PEERS: ${POOL_RPC_ROUTER_NODE_HEALTH_MIN_CONSENSUS_PEERS:-2}
-      POOL_RPC_ROUTER_EVM_HEAD_GUARD_ENABLED: ${POOL_RPC_ROUTER_EVM_HEAD_GUARD_ENABLED:-true}
+      POOL_RPC_ROUTER_EVM_HEAD_GUARD_ENABLED: ${POOL_RPC_ROUTER_EVM_HEAD_GUARD_ENABLED:-false}
       POOL_RPC_ROUTER_EVM_HEAD_GUARD_MAX_LAG_BLOCKS: ${POOL_RPC_ROUTER_EVM_HEAD_GUARD_MAX_LAG_BLOCKS:-30}
       POOL_RPC_ROUTER_EVM_REFERENCE_URLS: ${POOL_RPC_ROUTER_EVM_REFERENCE_URLS:-https://rpc.blockdag.engineering,https://rpc.bdagscan.com}
-      POOL_TEMPLATE_TTL_REFRESH_MS: ${POOL_TEMPLATE_TTL_REFRESH_MS:-2000}
+      POOL_TEMPLATE_TTL_REFRESH_MS: ${POOL_TEMPLATE_TTL_REFRESH_MS:-100}
       POOL_SUBMIT_STALE_BLOCK_CANDIDATES: ${POOL_SUBMIT_STALE_BLOCK_CANDIDATES:-false}
-      POOL_RECENT_STALE_BLOCK_CANDIDATE_SUBMIT_GRACE_MS: ${POOL_RECENT_STALE_BLOCK_CANDIDATE_SUBMIT_GRACE_MS:-250}
-      POOL_MAX_BLOCK_CANDIDATE_JOB_AGE_MS: ${POOL_MAX_BLOCK_CANDIDATE_JOB_AGE_MS:-2500}
+      POOL_RECENT_STALE_BLOCK_CANDIDATE_SUBMIT_GRACE_MS: ${POOL_RECENT_STALE_BLOCK_CANDIDATE_SUBMIT_GRACE_MS:-1750}
+      POOL_MAX_BLOCK_CANDIDATE_JOB_AGE_MS: ${POOL_MAX_BLOCK_CANDIDATE_JOB_AGE_MS:-1750}
       POOL_AUTO_TUNE_BLOCK_CANDIDATE_JOB_AGE: ${POOL_AUTO_TUNE_BLOCK_CANDIDATE_JOB_AGE:-true}
       POOL_AUTO_TUNE_BLOCK_CANDIDATE_MAX_AGE_MS: ${POOL_AUTO_TUNE_BLOCK_CANDIDATE_MAX_AGE_MS:-8000}
       POOL_SUBMIT_BLOCK_HEADER_V2_ENABLED: ${POOL_SUBMIT_BLOCK_HEADER_V2_ENABLED:-true}
       POOL_STALE_RACE_REJECT_WINDOW_SECONDS: ${POOL_STALE_RACE_REJECT_WINDOW_SECONDS:-10}
       POOL_STALE_RACE_CLIENT_RESEND_THRESHOLD: ${POOL_STALE_RACE_CLIENT_RESEND_THRESHOLD:-1}
-      POOL_STALE_RACE_CLIENT_RECONNECT_THRESHOLD: ${POOL_STALE_RACE_CLIENT_RECONNECT_THRESHOLD:-2}
+      POOL_STALE_RACE_CLIENT_RECONNECT_THRESHOLD: ${POOL_STALE_RACE_CLIENT_RECONNECT_THRESHOLD:-0}
       POOL_STALE_RACE_CLIENT_RECONNECT_WINDOW_SECONDS: ${POOL_STALE_RACE_CLIENT_RECONNECT_WINDOW_SECONDS:-120}
       POOL_STALE_RACE_CLIENT_RECONNECT_COOLDOWN_SECONDS: ${POOL_STALE_RACE_CLIENT_RECONNECT_COOLDOWN_SECONDS:-60}
       POOL_STALE_RACE_RECOVERY_COOLDOWN_SECONDS: ${POOL_STALE_RACE_RECOVERY_COOLDOWN_SECONDS:-5}
@@ -203,10 +215,10 @@ services:
       POOL_BLOCK_TIMING_CONTROLLER_WINDOW_SECONDS: ${POOL_BLOCK_TIMING_CONTROLLER_WINDOW_SECONDS:-60}
       POOL_BLOCK_TIMING_CONTROLLER_MIN_CANDIDATES: ${POOL_BLOCK_TIMING_CONTROLLER_MIN_CANDIDATES:-20}
       POOL_BLOCK_TIMING_CONTROLLER_TARGET_WASTE_PERCENT: ${POOL_BLOCK_TIMING_CONTROLLER_TARGET_WASTE_PERCENT:-5}
-      POOL_BLOCK_TIMING_CONTROLLER_MIN_JOB_AGE_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MIN_JOB_AGE_MS:-250}
-      POOL_BLOCK_TIMING_CONTROLLER_MAX_JOB_AGE_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MAX_JOB_AGE_MS:-2500}
+      POOL_BLOCK_TIMING_CONTROLLER_MIN_JOB_AGE_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MIN_JOB_AGE_MS:-1750}
+      POOL_BLOCK_TIMING_CONTROLLER_MAX_JOB_AGE_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MAX_JOB_AGE_MS:-1750}
       POOL_BLOCK_TIMING_CONTROLLER_MIN_TEMPLATE_TTL_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MIN_TEMPLATE_TTL_MS:-100}
-      POOL_BLOCK_TIMING_CONTROLLER_MAX_TEMPLATE_TTL_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MAX_TEMPLATE_TTL_MS:-1000}
+      POOL_BLOCK_TIMING_CONTROLLER_MAX_TEMPLATE_TTL_MS: ${POOL_BLOCK_TIMING_CONTROLLER_MAX_TEMPLATE_TTL_MS:-100}
       POOL_BLOCK_TIMING_CONTROLLER_STEP_MS: ${POOL_BLOCK_TIMING_CONTROLLER_STEP_MS:-250}
       POOL_BLOCK_TIMING_CONTROLLER_COOLDOWN_SECONDS: ${POOL_BLOCK_TIMING_CONTROLLER_COOLDOWN_SECONDS:-20}
       BDAG_ASIC_LAN_CIDRS: ${BDAG_ASIC_LAN_CIDRS:-}
@@ -272,6 +284,7 @@ services:
       NODE_RPC_USER: ${NODE_RPC_USER:-test}
       NODE_RPC_PASS: ${NODE_RPC_PASS:-test}
       BDAG_WATCHDOG_INTERVAL: ${BDAG_WATCHDOG_INTERVAL:-60}
+      BDAG_WATCHDOG_NATIVE_P2P_PEER_LOSS_REPAIR_SECONDS: ${BDAG_WATCHDOG_NATIVE_P2P_PEER_LOSS_REPAIR_SECONDS:-300}
     volumes:
       - /var/run/docker.sock:/var/run/docker.sock
       - ${BDAG_STACK_HOST_ROOT:-${PWD}}:${BDAG_STACK_HOST_ROOT:-${PWD}}
@@ -431,6 +444,7 @@ services:
       - /run/dashboard-redis:size=16m,mode=0700
     volumes:
       - dashboard-redis:/var/lib/dashboard-redis
+      - ${BDAG_HOST_ARP_TABLE_PATH:-/proc/net/arp}:/host/proc/net/arp:ro
     depends_on:
       node: { condition: service_started }
       pool: { condition: service_started }
@@ -451,12 +465,13 @@ services:
       BDAG_REDIS_MAXMEMORY_POLICY: ${BDAG_REDIS_MAXMEMORY_POLICY:-noeviction}
       BDAG_EVM_HTTP_URL: ${BDAG_EVM_HTTP_URL:-http://host.docker.internal:18545}
       BDAG_EVM_WS_URL: ${BDAG_EVM_WS_URL:-ws://host.docker.internal:18546}
+      BDAG_EVM_SYNC_BACKOFF_SECONDS: ${BDAG_EVM_SYNC_BACKOFF_SECONDS:-60}
       BDAG_POOL_METRICS_URL: ${BDAG_POOL_METRICS_URL:-http://host.docker.internal:9090/metrics}
       BDAG_NODE_RPC_URL: ${BDAG_NODE_RPC_URL:-http://host.docker.internal:38131}
       BDAG_GLOBAL_BLOCK_WINDOW: ${BDAG_GLOBAL_BLOCK_WINDOW:-600}
       BDAG_GLOBAL_CACHE_MAX_TIP_LAG_BLOCKS: ${BDAG_GLOBAL_CACHE_MAX_TIP_LAG_BLOCKS:-30}
       BDAG_GLOBAL_FAST_WARMUP_BLOCKS: ${BDAG_GLOBAL_FAST_WARMUP_BLOCKS:-120}
-      BDAG_GLOBAL_RPC_WORKERS: ${BDAG_GLOBAL_RPC_WORKERS:-24}
+      BDAG_GLOBAL_RPC_WORKERS: ${BDAG_GLOBAL_RPC_WORKERS:-4}
       BDAG_GLOBAL_SETTLE_DELAY_SECONDS: ${BDAG_GLOBAL_SETTLE_DELAY_SECONDS:-5}
       BDAG_GLOBAL_MAX_INCREMENTAL_GAP: ${BDAG_GLOBAL_MAX_INCREMENTAL_GAP:-120}
       BDAG_GLOBAL_RPC_CHUNK_SIZE: ${BDAG_GLOBAL_RPC_CHUNK_SIZE:-64}
@@ -478,7 +493,9 @@ services:
       DASHBOARD_TIMEZONE: ${DASHBOARD_TIMEZONE:-Africa/Johannesburg}
       BDAG_MINER_CONTROL_ENABLED: ${BDAG_MINER_CONTROL_ENABLED:-1}
       BDAG_ASIC_LAN_CIDRS: ${BDAG_ASIC_LAN_CIDRS:-}
+      POOL_ASIC_ARP_TABLE_PATH: ${POOL_ASIC_ARP_TABLE_PATH:-/host/proc/net/arp}
       POOL_ASIC_MAC_OVERRIDES: ${POOL_ASIC_MAC_OVERRIDES:-}
+      POOL_ASIC_MAC_ALLOWLIST: ${POOL_ASIC_MAC_ALLOWLIST:-}
       BDAG_POOL_URL: ${BDAG_POOL_URL:-}
       MINING_POOL_ADDRESS: ${MINING_POOL_ADDRESS:?set MINING_POOL_ADDRESS to a non-zero payout address}
       POOL_COINBASE_ADDRESS: ${POOL_COINBASE_ADDRESS:-}