外部 Redis "Queue" 模式 "可能是 n8n 的 bug?"

問題
,
1

n8n 佇列模式：從叢集內 Redis 切換到外部 Aiven Valkey 後出現 PostgreSQL 連線逾時

描述問題/錯誤/疑問

摘要
在將 Redis 從叢集內部署遷移到外部 Aiven Valkey 後，n8n 開始出現 PostgreSQL 連線逾時的執行失敗。
在 Redis 遷移之前，相同的部署運作正常。（任務未被拾取但可運作）

環境

  • 透過 Helm chart 部署的 n8n
  • Kubernetes 叢集
  • 已啟用佇列模式
  • PostgreSQL 在叢集內執行
  • Redis 已從叢集內 Redis 遷移到外部 Aiven Valkey

anton% kubectl get pods -n n8n
NAME READY STATUS RESTARTS AGE
n8n-77889b4756-9wqtq 1/1 Running 0 9m6s
n8n-postgresql-0 1/1 Running 0 112m
n8n-redis-master-0 1/1 Running 0 9m6s
n8n-webhook-7ffccd5dd8-njpnm 1/1 Running 1 (8m48s ago) 9m6s
n8n-worker-f7fff89fb-k45ff 1/1 Running 0 8m20s
n8n-worker-f7fff89fb-nv994 1/1 Running 1 (8m42s ago) 9m6s

錯誤訊息是什麼（如有的話）？

請分享你的工作流程

Lost Redis connection. Trying to reconnect in 1s...

Recovered Redis connection
Execution XX (job YY) failed

Error: timeout exceeded when trying to connect

簡單 Webhook

image
image482×316 10.6 KB


Error: timeout exceeded when trying to connect
    at pg-pool/index.js:45
    at PostgresDriver.obtainMasterConnection
    at PostgresQueryRunner.query
    at ExecutionRepository.findSingleExecution
    at JobProcessor.processJob

變更內容

唯一的基礎設施變更是將 Redis 從叢集內 Redis 執行個體遷移到外部 Aiven Valkey。

遷移前：

  • 佇列模式運作正常
  • 執行成功完成
  • 沒有 PostgreSQL 逾時錯誤

遷移後：

  • Redis 重新連線訊息開始出現

  • 工作程式執行失敗

  • 出現 PostgreSQL 連線逾時錯誤

    其他觀察

    從工作程式 Pod 到 Aiven Valkey 的連線狀況良好：

    23 packets transmitted

23 received

0% packet loss

avg latency ~16ms

使用外部 Redis

image
image1225×713 61.6 KB

使用叢集內 Redis

image
image1331×759 76.3 KB

關於你的 n8n 設定的資訊

  • n8n 版本： 2.25.7
  • 資料庫（預設：postgresql）：
  • n8n EXECUTIONS_PROCESS 設定（預設：own, main）：
  • 執行 n8n 的方式 (kuber)：
  • 作業系統： Debian12
    我在最新版本 v1 上試過 問題相同
2

嗨 @Antonio1

Aiven Valkey 需要特定的配置，這些配置與標準的集群內 Redis 部署不同。

  • TLS/SSL 要求： Aiven 服務需要 TLS。確保你的 n8n 環境變數包含：
    • N8N_REDIS_TLS=true（或根據你的 n8n 版本/自訂映像的等效設定）。
    • 如果你使用自訂連接字串，請確保它使用 rediss://（兩個 ‘s’）而不是 redis://。
  • TCP 保活： 外部連接更容易被雲防火牆/負載平衡器中斷。確保你的 Kubernetes 節點具有適當的 TCP 保活設定，或檢查 Aiven 是否提供建議的保活間隔。
  • 網路政策： 確認你的 Kubernetes 出口政策允許在 Valkey 連接埠（通常為 6379 或自訂 Aiven 連接埠）上到外部 Aiven IP 範圍的流量。

請告訴我這些是否有幫助

3

感謝你的回覆

是的，我知道我試過了

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT: 30000
QUEUE_BULL_REDIS_TIMEOUT: 30000
QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY: 15000

而且是的，我正在使用 Aiven 的 TLS

tls:
    enabled: true
host: valkey-XXXXXXX.l.aivencloud.com
port: 21275

username: default
password: "PASS_HERE"

n8N 的日誌

Execution 29 (job 99) failed

Error: timeout exceeded when trying to connect

    at /usr/local/lib/node_modules/n8n/node_modules/.pnpm/pg-pool@3.6.2_pg@8.12.0/node_modules/pg-pool/index.js:45:11

    at PostgresDriver.obtainMasterConnection (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresDriver.ts:1181:28)

    at PostgresQueryRunner.query (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresQueryRunner.ts:248:36)

    at SelectQueryBuilder.loadRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3660:25)

    at SelectQueryBuilder.getRawMany (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1617:29)

    at SelectQueryBuilder.executeEntitiesAndRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3322:26)

    at SelectQueryBuilder.getRawAndEntities (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1661:29)

    at SelectQueryBuilder.getOne (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1688:25)

    at ExecutionRepository.findSingleExecution (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+db@file+packages+@n8n+db_@opentelemetry+api@1.9.0_@opentelemetry+sdk-trace-base@1._ab22bba05a964211b9fe14bf4b841570/node_modules/@n8n/db/src/repositories/execution.repository.ts:315:21)

    at JobProcessor.processJob (/usr/local/lib/node_modules/n8n/src/scaling/job-processor.ts:57:21)

執行 29 出現問題：錯誤：嘗試連線時逾時。正在中止。

Error: timeout exceeded when trying to connect (execution 29)

Error: Error: timeout exceeded when trying to connect

    at Queue.onFailed (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/bull@4.16.4_patch_hash=a4b6d56db16fe5870646929938466d6a5c668435fd1551bed6a93fffb597ba42/node_modules/bull/lib/job.js:523:18)

    at processTicksAndRejections (node:internal/process/task_queues:104:5)

Execution 30 (job 100) failed

Error: timeout exceeded when trying to connect

    at /usr/local/lib/node_modules/n8n/node_modules/.pnpm/pg-pool@3.11.0_pg@8.17.0/node_modules/pg-pool/index.js:45:11

    at PostgresDriver.obtainMasterConnection (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresDriver.ts:1182:28)

    at PostgresQueryRunner.query (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresQueryRunner.ts:248:36)

    at SelectQueryBuilder.loadRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3660:25)

    at SelectQueryBuilder.getRawMany (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1617:29)

    at SelectQueryBuilder.executeEntitiesAndRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3322:26)

    at SelectQueryBuilder.getRawAndEntities (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1661:29)

    at SelectQueryBuilder.getOne (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-16_@sentry+node@10.36.0_mysql2@3.17.0_pg@8.17.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1688:25)
4

可能嗎？

Job 在 Redis 中卡住 ↓
Main n8n 看到執行狀態為「running」↓
嘗試從 PostgreSQL 讀取以顯示狀態 ↓
pg-pool 超時——因為許多這樣的「掛起」請求同時堆積

5

是的，那個形狀是可能的，但你貼的追蹤資料是 pg-pool，所以下一個瓶頸是 Postgres 在 worker 並行下的容量，而不是另一個 Redis 逾時設定。幾個卡住的任務可以讓執行讀操作保持開啟的時間足夠長，使得 main/webhook/workers 都觸及同一個資料庫上限。

有個小檢查：有多少個 workers 在運行，DB_POSTGRESDB_POOL_SIZE 是多少，Postgres max_connections 在保留位置之後是多少？如果連接池的數學運算已經很緊張，Valkey 就暴露了資料庫的瓶頸。

6

@Antonio1

你需要為資料庫連線提供更多「呼吸空間」，這樣即使 worker 行為不當，狀態檢查也能成功。

  • 在你的 n8n 部署中設定以下環境變數：DB_POSTGRESDB_POOL_SIZE: "50"（或更高，取決於你的總 worker 數量）。
  • 關鍵： 確保你的 PostgreSQL max_connections（在 postgresql.conf 或你的雲端提供商設定中）至少設定為 (pod_數量 * DB_POSTGRESDB_POOL_SIZE) + 20。

由於你使用的是 Aiven，在重新連線事件期間 TLS 握手可能是瓶頸。

  • 增加 Redis 連線逾時： 你已將 QUEUE_BULL_REDIS_CONNECTION_TIMEOUT 設定為 30 秒。試著將其增加至 60000（60 秒），以便在公網上進行較慢的 TLS 握手。
  • 停用 Redis 叢集模式（如適用）： 如果你沒有明確使用 Valkey/Redis 叢集，請確保 n8n 設定為單一節點連線。
  • 檢查 Aiven 指標： 登入你的 Aiven 控制台，檢查你 Valkey 服務的**「連線數量」和「CPU 使用率」**。如果在你的 n8n worker 失敗的時間內看到尖峰，你可能已達到 Aiven 端的連線限制。
7

好的，我剛剛再測試了一次，只用了一個 POST

在執行標籤中，這個 POST 顯示為 佇列 然後 失敗
這個 n8n 上沒有其他工作流程，我只是在這裡試著使用佇列模式
這個工作沒有出現在任何工作程序上
它只是在主要 N8n Pod 上失敗

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection

Enqueued execution 122 (job 166)

Execution 122 (job 166) failed

Error: timeout exceeded when trying to connect

    at /usr/local/lib/node_modules/n8n/node_modules/.pnpm/pg-pool@3.6.2_pg@8.12.0/node_modules/pg-pool/index.js:45:11

    at PostgresDriver.obtainMasterConnection (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresDriver.ts:1181:28)

    at PostgresQueryRunner.query (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresQueryRunner.ts:248:36)

    at SelectQueryBuilder.loadRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3660:25)

    at SelectQueryBuilder.getRawMany (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1617:29)

    at SelectQueryBuilder.executeEntitiesAndRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3322:26)

    at SelectQueryBuilder.getRawAndEntities (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1661:29)

    at SelectQueryBuilder.getOne (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1688:25)

    at ExecutionRepository.findSingleExecution (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+db@file+packages+@n8n+db_@opentelemetry+api@1.9.0_@opentelemetry+sdk-trace-base@1._ab22bba05a964211b9fe14bf4b841570/node_modules/@n8n/db/src/repositories/execution.repository.ts:315:21)

    at JobProcessor.processJob (/usr/local/lib/node_modules/n8n/src/scaling/job-processor.ts:57:21)

Problem with execution 122: Error: timeout exceeded when trying to connect. Aborting.

Error: timeout exceeded when trying to connect (execution 122)

Error: Error: timeout exceeded when trying to connect

    at Queue.onFailed (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/bull@4.16.4_patch_hash=a4b6d56db16fe5870646929938466d6a5c668435fd1551bed6a93fffb597ba42/node_modules/bull/lib/job.js:523:18)

    at processTicksAndRejections (node:internal/process/task_queues:104:5)

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection
8

這改變了讀取方式。如果一次測試執行從未在任何 worker 上執行，這還不是 Postgres pool-size 問題。該工作被排入隊列，但 worker 端未能保持足夠長的時間連接到相同的 Redis/Valkey 隊列以領取它。

發佈一條 worker 啟動日誌行，其中顯示它以隊列模式連接，以及 EXECUTIONS_MODE、QUEUE_BULL_REDIS_HOST/PORT/TLS 和 QUEUE_BULL_PREFIX（移除機密）。首先要確認的是主程序和 worker 指向的是完全相同的隊列配置，而不只是相同的 Valkey 主機。

9

@Antonio1 :white_check_mark: 在 n8n 工作流程中管理狀態指標和記憶體問題需要清潔、模組化的架構，特別是在跨 Google Docs 和 Sheets 同步複雜資料時。

:white_check_mark: 如果您想簡化此邏輯或需要強大的生產就緒解決方案來穩定您的 AI 代理，讓我們聯繫。

:white_check_mark: 我專門構建和優化高效能的 n8n、AI、Make.com 和 GHL 自動化系統

10 
EXECUTIONS_MODE

{26-06-12 12:52}Antons-MacBook-Pro-2:~ anton% for pod in $(kubectl get pods -n n8n -o name); do
echo "=== $pod ==="
kubectl exec -n n8n ${pod#pod/} -- env | grep EXECUTIONS_MODE
done
=== pod/n8n-564fbc565f-lxnsc ===
已預設容器「n8n-main」，共有：n8n-main、npm-install (init)
EXECUTIONS_MODE=queue
=== pod/n8n-postgresql-0 ===
=== pod/n8n-webhook-6bfc9f5cd8-dcv7l ===
EXECUTIONS_MODE=queue
=== pod/n8n-worker-7fb4c7d87d-jzwth ===
已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)
EXECUTIONS_MODE=queue
=== pod/n8n-worker-7fb4c7d87d-k82xw ===
已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)
EXECUTIONS_MODE=queue
=== pod/n8n-worker-7fb4c7d87d-kkdm9 ===
已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)
EXECUTIONS_MODE=queue
=== pod/n8n-worker-7fb4c7d87d-lqt7j ===
已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)
EXECUTIONS_MODE=queue
=== pod/n8n-worker-7fb4c7d87d-mspxz ===
已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)
EXECUTIONS_MODE=queue
=== pod/n8n-worker-7fb4c7d87d-xqcr2 ===
已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)
EXECUTIONS_MODE=queue

且

{26-06-12 12:52}Antons-MacBook-Pro-2:~ anton% for pod in $(kubectl get pods -n n8n -o name); do

  echo "=== $pod ==="

  kubectl exec -n n8n ${pod#pod/} -- env | grep QUEUE_BULL_

done

=== pod/n8n-564fbc565f-lxnsc ===

已預設容器「n8n-main」，共有：n8n-main、npm-install (init)

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_REDIS_TLS=true

=== pod/n8n-postgresql-0 ===

=== pod/n8n-webhook-6bfc9f5cd8-dcv7l ===

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

=== pod/n8n-worker-7fb4c7d87d-jzwth ===

已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

=== pod/n8n-worker-7fb4c7d87d-k82xw ===

已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

=== pod/n8n-worker-7fb4c7d87d-kkdm9 ===

已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

=== pod/n8n-worker-7fb4c7d87d-lqt7j ===

已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_REDIS_PORT=21275

=== pod/n8n-worker-7fb4c7d87d-mspxz ===

已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_REDIS_USERNAME=default

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

=== pod/n8n-worker-7fb4c7d87d-xqcr2 ===

已預設容器「n8n-worker」，共有：n8n-worker、npm-install (init)

QUEUE_BULL_REDIS_DUALSTACK=false

QUEUE_BULL_REDIS_KEEP_ALIVE_DELAY=15000

QUEUE_BULL_REDIS_TLS=true

QUEUE_BULL_REDIS_TIMEOUT=30000

QUEUE_BULL_REDIS_DB=0

QUEUE_BULL_SETTINGS_STALLED_INTERVAL=60000

QUEUE_BULL_REDIS_PASSWORD=PASSWORD_SECURE

QUEUE_BULL_SETTINGS_LOCK_DURATION=60000

QUEUE_BULL_SETTINGS_LOCK_RENEW_TIME=30000

QUEUE_BULL_REDIS_HOST=valkey-757370-XXXX.l.aivencloud.com

QUEUE_BULL_REDIS_PORT=21275

QUEUE_BULL_REDIS_CONNECTION_TIMEOUT=30000

QUEUE_BULL_REDIS_USERNAME=default
11

那些 env dumps 很有用：main、webhook 和 workers 現在都指向同一個 Valkey 佇列，所以這看起來不再像是 pods 之間有不同的 Redis 主機/連接埠/TLS 設定的問題。

那兩行遺漏的日誌是 worker 側實際啟動/聲稱工作的部分和 DB 連接池上限。貼一份 worker 在啟動/job 166 左右的日誌，加上 DB_POSTGRESDB_POOL_SIZE 和 Postgres 的 max_connections；有六個 workers 時，即使 Redis 設定相符，較小的 Postgres 連接池/max_connections 仍然可能導致故障呈現為 pg-pool。

12

{26-06-12 15:07}Antons-MacBook-Pro-2:~ anton% kubectl get pods -n n8n
NAME READY STATUS RESTARTS AGE
n8n-6687b7fb7d-vj2xr 1/1 Running 0 42s
n8n-postgresql-0 1/1 Running 0 17h
n8n-worker-7c5c5d5dc-x2x9d 1/1 Running 0 42s

好的，所以我把它簡化得很簡單
一個主節點 + 一個worker（postgresql 在集群中 ）

=Worker=
與 job 124 無關

Concurrency is set to less than 5. THIS CAN LEAD TO AN UNSTABLE ENVIRONMENT. Please consider increasing it to at least 5 to make best use of the worker.

n8n Task Broker ready on 127.0.0.1, port 5679

Failed to start Python task runner in internal mode. because Python 3 is missing from this system. Launching a Python runner in internal mode is intended only for debugging and is not recommended for production. Users are encouraged to deploy in external mode. See: 


[license SDK] Skipping renewal on init: renewOnInit is disabled in config

[license SDK] Skipping renewal on init: autoRenewEnabled is disabled in config

[license SDK] Skipping renewal on init: license cert is not initialized

Instance registered

Discovered 4 cluster checks

n8n worker server listening on port 5678

n8n worker is now ready

 * Version: 2.25.7

 * Concurrency: 2

Registered runner "JS Task Runner" (oqClB5JFV1-IPGaVNCRFu)

Worker started execution 123 (job 167)

Worker finished execution 123 (job 167)

=主節點=
execution 124

Defaulted container "n8n-main" out of: n8n-main, npm-install (init)

Initializing n8n process

n8n ready on ::, port 5678

[license SDK] Skipping renewal on init: license cert is not initialized

Attempting to reinstall missing packages

Failed to execute npm command

Failed to reinstall community package n8n-nodes-minio: Failed to execute npm command

Failed to execute npm command

Failed to reinstall community package n8n-nodes-firebird: Failed to execute npm command

Instance registered

Discovered 4 cluster checks

Cluster check warning Detected 2 instances claiming leader role: 0490eef8-9207-4ef7-a379-4699f18e001c, 2586c901-fd3c-4c06-a190-a4197e69f325

Currently active workflows:

- My workflow (ID: QFQvCXo8mBhadrAM)

Version: 2.25.7

Building workflow dependency index...

Start Active Workflows:

Finished building workflow dependency index. Processed 0 draft workflows, 0 published workflows.

Activated workflow "My workflow" (ID: QFQvCXo8mBhadrAM)




Editor is now accessible via:

URL-HERE

Enqueued execution 123 (job 167)

Execution 123 (job 167) finished

Enqueued execution 124 (job 168)

User attempted to access a workflow without permissions

Execution 124 (job 168) failed

Error: timeout exceeded when trying to connect

    at /usr/local/lib/node_modules/n8n/node_modules/.pnpm/pg-pool@3.6.2_pg@8.12.0/node_modules/pg-pool/index.js:45:11

    at PostgresDriver.obtainMasterConnection (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresDriver.ts:1181:28)

    at PostgresQueryRunner.query (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/driver/postgres/PostgresQueryRunner.ts:248:36)

    at SelectQueryBuilder.loadRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3660:25)

    at SelectQueryBuilder.getRawMany (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1617:29)

    at SelectQueryBuilder.executeEntitiesAndRawResults (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:3322:26)

    at SelectQueryBuilder.getRawAndEntities (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1661:29)

    at SelectQueryBuilder.getOne (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+typeorm@0.3.20-15_@sentry+node@9.42.1_mysql2@3.15.0_pg@8.12.0_sqlite3@5.1.7/node_modules/src/query-builder/SelectQueryBuilder.ts:1688:25)

    at ExecutionRepository.findSingleExecution (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/@n8n+db@file+packages+@n8n+db_@opentelemetry+api@1.9.0_@opentelemetry+sdk-trace-base@1._ab22bba05a964211b9fe14bf4b841570/node_modules/@n8n/db/src/repositories/execution.repository.ts:315:21)

    at JobProcessor.processJob (/usr/local/lib/node_modules/n8n/src/scaling/job-processor.ts:57:21)




Problem with execution 124: Error: timeout exceeded when trying to connect. Aborting.

Error: timeout exceeded when trying to connect (execution 124)

Error: Error: timeout exceeded when trying to connect

    at Queue.onFailed (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/bull@4.16.4_patch_hash=a4b6d56db16fe5870646929938466d6a5c668435fd1551bed6a93fffb597ba42/node_modules/bull/lib/job.js:523:18)

    at processTicksAndRejections (node:internal/process/task_queues:104:5)




User attempted to access a workflow without permissions

User attempted to access a workflow without permissions

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection

Lost Redis connection. Trying to reconnect in 1s... (0s/10s)

Recovered Redis connection

image
image1676×852 113 KB

13

125 OK
所以這是某種隨機的

image
image1478×827 84.5 KB

==Worker==

n8n worker is now ready

Version: 2.25.7

Concurrency: 2

Registered runner "JS Task Runner" (oqClB5JFV1-IPGaVNCRFu)
Worker started execution 123 (job 167)
Worker finished execution 123 (job 167)
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Worker started execution 125 (job 169)
Worker finished execution 125 (job 169)

=Main=

Problem with execution 124: Error: timeout exceeded when trying to connect. Aborting.
Error: timeout exceeded when trying to connect (execution 124)
Error: Error: timeout exceeded when trying to connect
at Queue.onFailed (/usr/local/lib/node_modules/n8n/node_modules/.pnpm/bull@4.16.4_patch_hash=a4b6d56db16fe5870646929938466d6a5c668435fd1551bed6a93fffb597ba42/node_modules/bull/lib/job.js:523:18)
at processTicksAndRejections (node:internal/process/task_queues:104:5)

User attempted to access a workflow without permissions
User attempted to access a workflow without permissions
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Lost Redis connection. Trying to reconnect in 1s... (0s/10s)
Recovered Redis connection
Enqueued execution 125 (job 169)
Execution 125 (job 169) finished
User attempted to access a workflow without permissions
14

我進行了額外的測試
我在 n8n 工作區內啟動了一個新的 pod

redis-cli
-h ``valkey-XXXX.l.aivencloud.com``
-p 21275
--tls
--user default
-a 'YOUR_PASSWORD'
SUBSCRIBE test

然後使用我電腦上的另一個 CLIENT 來發送事件

while true; do
redis-cli
-h ``valkey-XXXX.l.aivencloud.com``
-p 21275
--tls
--user default
-a 'YOUR_PASSWORD'
PUBLISH test hello
sleep 1
done

=結果=
新 pod 未中斷與 redis 的連線

  1. “hello”

  2. “message”

  3. “test”

  4. “hello”

  5. “message”

  6. “test”

  7. “hello”

  8. “message”

  9. “test”

    =Worker n8n 方面=

    
Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)

Redis 連線已恢復

Redis 連線已中斷。正在嘗試在 1 秒內重新連線... (0 秒/10 秒)