Airflow 1.10.15 : The CSRF session token is missing when i try to trigger a new dag

Apache Airflow version: 1.10.15 https://raw.githubusercontent.com/apache/airflow/constraints-1.10.15/constraints-3.6.txt

Kubernetes version: Client Version: v1.16.2 Server Version: v1.14.8-docker-1

Environment: python 3.6.8 + celeryExecutor + rbac set to false

• OS (e.g. from /etc/os-release): CentOS Linux 7 (Core)
• Kernel (e.g. uname -a): 3.10.0-1127.19.1.el7.x86_64

What happened: I have upgraded from 1.10.12 to 1.10.15, when i trigger a dag i have the exception below

What you expected to happen: trigger a dag without exceptions

How to reproduce it: use airflow 1.10.15 and try to trigger an example dag example_bash_operator

Anything else we need to know:

How often does this problem occur: Every time i trigger a dag

Any relevant logs to include? Put them here in side a detail tag:

webserver.log

[2021-04-26 15:03:06,611] {__init__.py:50} INFO - Using executor CeleryExecutor [2021-04-26 15:03:06,612] {dagbag.py:417} INFO - Filling up the DagBag from /home/airflow/dags 175.62.58.93 - - [26/Apr/2021:15:03:10 +0000] "GET /health HTTP/1.1" 200 187 "-" "kube-probe/1.14+" 175.62.58.93 - - [26/Apr/2021:15:03:11 +0000] "GET /health HTTP/1.1" 200 187 "-" "kube-probe/1.14+" 175.62.58.93 - - [26/Apr/2021:15:03:15 +0000] "GET /health HTTP/1.1" 200 187 "-" "kube-probe/1.14+" 175.62.58.93 - - [26/Apr/2021:15:03:16 +0000] "GET /health HTTP/1.1" 200 187 "-" "kube-probe/1.14+" [2021-04-26 15:03:17,401] {csrf.py:258} INFO - The CSRF session token is missing. 10.30.180.137 - - [26/Apr/2021:15:03:17 +0000] "POST /admin/airflow/trigger?dag_id=example_bash_operator&origin=https://xxxxxx/admin/ HTTP/1.1" 400 150 "https://xxxxxxxxxxxx/admin/airflow/trigger?dag_id=example_bash_operator&origin=https://xxxxxxxxxxxx/admin/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.183 Safari/537.36" 175.62.58.93 - - [26/Apr/2021:15:03:20 +0000] "GET /health HTTP/1.1" 200 187 "-" "kube-probe/1.14+" 175.62.58.93 - - [26/Apr/2021:15:03:21 +0000] "GET /health HTTP/1.1" 200 187 "-" "kube-probe/1.14+"

[kaxil]

@chakir-alaoui When you use more than 1 Webservers, make sure [webserver] secret_key is set same across both of them.

If you don’t set it by it will generate a random ones and since both webserver will have different keys, one of them will error.

[chakir-alaoui]

thank you for your help, it’s working now 👍