bug: the version 3.14.1 or 3.14.0 get init_worker_by_lua error: /usr/local/apisix//deps/share/lua/5.1/resty/http.lua:74: API disabled in the context of init_worker_by_lua

[shijunLee]

Current Behavior

get init_worker_by_lua error: /usr/local/apisix//deps/share/lua/5.1/resty/http.lua:74: API disabled in the context of init_worker_by_lua* error

Expected Behavior

fix this bug

Error Logs

2025/10/29 09:45:27 [warn] 25958#25958: 17772 [lua] config_etcd.lua:1208: init_worker(): master process has been running for a long time, reloading the full configuration from etcd for this new worker, context: init_worker_by_lua
2025/10/29 09:45:27 [error] 25958#25958: init_worker_by_lua error: /usr/local/apisix//deps/share/lua/5.1/resty/http.lua:74: API disabled in the context of init_worker_by_lua*
stack traceback:
[C]: in function 'co_create'
/usr/local/apisix//deps/share/lua/5.1/resty/http.lua:74: in function '_body_reader'
/usr/local/apisix//deps/share/lua/5.1/resty/http.lua:821: in function 'request'
/usr/local/apisix//deps/share/lua/5.1/resty/etcd/v3.lua:120: in function 'request_uri_via_unix_socket'
/usr/local/apisix//deps/share/lua/5.1/resty/etcd/v3.lua:160: in function 'http_request_uri'
/usr/local/apisix//deps/share/lua/5.1/resty/etcd/v3.lua:251: in function '_request_uri'
/usr/local/apisix//deps/share/lua/5.1/resty/etcd/v3.lua:535: in function 'refresh_jwt_token'
/usr/local/apisix//deps/share/lua/5.1/resty/etcd/v3.lua:205: in function '_request_uri'
/usr/local/apisix//deps/share/lua/5.1/resty/etcd/v3.lua:670: in function 'readdir'
/usr/local/apisix/apisix/core/config_etcd.lua:404: in function 'readdir'
/usr/local/apisix/apisix/core/config_etcd.lua:1167: in function 'init_loaded_configuration'
/usr/local/apisix/apisix/core/config_etcd.lua:1210: in function 'init_worker'
/usr/local/apisix/apisix/init.lua:137: in function 'http_init_worker'
init_worker_by_lua:2: in main chunk

Steps to Reproduce

1. run apisix with docker ,the docker image is pull form docker hub
2. wait time and exec this job

Environment

• APISIX version (run apisix version): 3.14.1
• Operating system (run uname -a): Debian GNU/Linux 11 (bullseye)
• OpenResty / Nginx version (run openresty -V or nginx -V): nginx version: openresty/1.27.1.2
• etcd version, if relevant (run curl http://127.0.0.1:9090/v1/server_info):
• APISIX Dashboard version, if relevant:
• Plugin runner version, for issues related to plugin runners:
• LuaRocks version, for installation issues (run luarocks --version):

[shijunLee]

i found this pr api7/apisix-nginx-module#108 but the office apisix images use the openresty/1.27.1.2 , is this not support

[Baoyuantop]

Hi @shijunLee, your Steps to Reproduce description is too brief. Could you elaborate on how this issue is triggered? Also, do you have any custom plugins or modifications to the APISIX code?

[shijunLee]

Hi @shijunLee, your Steps to Reproduce description is too brief. Could you elaborate on how this issue is triggered? Also, do you have any custom plugins or modifications to the APISIX code?

@Baoyuantop i develop some custom plugins and not change the apisix code ,only add plugins , this bug is the nginx module this api7/apisix-nginx-module#108 change the nginx model code enable co_create call in nginx init_worker_by_lua ,but i do not know the default docker hub apisix images ，is use the apisix nginx module or not . the apisix image use the openresty version is 1.27.1.2,but the apisix nginx module is only have 1.27.1.1 patch. i modfy the etcd_config.lua , disable the full data reload ,disable the code with comments，code is

if configuration_loaded_time and ngx_time() - configuration_loaded_time > threshold then
        log.warn("master process has been running for a long time, ",
                     "reloading the full configuration from etcd for this new worker")
        local err = init_loaded_configuration()
        if err then
            return nil, err
        end
    end