apisix-helm-chart
apisix-helm-chart copied to clipboard
怎么调整 lua_shared_dict prometheus-metrics
helm install apisix
chart version : apisix-0.11.0
values.yaml apisix: customLuaSharedDicts:
- name: prometheus-metrics size: 200m
错误信息: 2022/09/21 13:25:01 [emerg] 1#1: the size 209715200 of shared memory zone "prometheus-metrics" conflicts with already declared size 10485760 in /usr/local/apisix/conf/nginx.conf:147 nginx: [emerg] the size 209715200 of shared memory zone "prometheus-metrics" conflicts with already declared size 10485760 in /usr/local/apisix/conf/nginx.conf:147
You cannot tweak the shared memory size and reload APISIX. This is due to the Nginx limitation.
The way to change the size is "hot reload" APISIX, but it's not easy on Kubernetes. Doe this cluster already handle real traffic?
Yes, it's already real traffic
Wait, how did you execute the reload? In Kubernetes, we don't use "reload" but just create new pods to replace old ones.
helm upgrade -install apisix apisix/apisix
--values apisix.nginx.conf.yaml
apisix.nginx.conf.yaml
apisix: customLuaSharedDicts: - name: prometheus-metrics size: 200m
I've got the same problem.