worker 1 (pid: 150) died, killed by signal 9 :( trying respawn

[hjing100]

nebula：2.6.0
使用uwsgi+Django+docker

问题：当下查库都没有问题，过了几个小时再查，就报错，下一个问题查库就会自动重新连接nebula，给出正确答案

报错：
Thu Mar 10 20:44:36 2022 - *** HARAKIRI ON WORKER 1 (pid: 150, try: 1) ***
Thu Mar 10 20:44:36 2022 - HARAKIRI !!! worker 1 status !!!
Thu Mar 10 20:44:36 2022 - HARAKIRI [core 2] 172.17.0.1 - POST /kbqa since 1646916215
Thu Mar 10 20:44:36 2022 - HARAKIRI !!! end of worker 1 status !!!
DAMN ! worker 1 (pid: 150) died, killed by signal 9 :( trying respawn ...

[hjing100]

一次连接会话持续 30 分钟。如果超过 30 分钟没有操作，会话即断开，用户需要重新登录数据库。
请问这个需要在哪里设置，改善这种现象

[Aiee]

As you stated in the description, the signal was sent to HARAKIRI, it seems you need to check the timeout with uwsgi, and the Nebula service seems to not be involved with the information you provided.

And just a note to you, you could change session_idle_timeout_secs in the conf file to modify the session timeout of the Nebula service.

[wey-gu]

ref: https://discuss.nebula-graph.com.cn/t/topic/7864