Worker process exits on signal 11 when trying to purge


#1

Hi, I have NGINX compiled with the --add-module=/tmp/naxsi/naxsi_src --add-module=/tmp/ngx_cache_purge module options (placed first in the ./configure order, if that matters). All of this stuff is running in Docker on a Kubernetes cluster.

I’m on a single site setup using this configuration. I have the fastcgi_cache_path, fastcgi_cache_key, fastcgi_cache_use_stale, and fastcgi_ignore_headers directives in the http {} context.

I can’t get any of the cache files to purge using either the “Purge Entire Cache” button or the PURGE/url method. Neither does the automatic purging work (I’m verifying by watching the /var/run/nginx-cache directory).

When I try the PURGE/url method I get an NGINX error in the log - [alert] 1#1: worker process 114 exited on signal 11.

Does the /var/run/nginx-cache directory have to be writable by the PHP FPM user? Currently that dir is with user/group nginx permissions and the FPM processes are user/group www-data.

Additionally, the NGINX Helper logger is just returning Purged Everything!. Not sure what else I can check/try at this point…

NGINX version:

nginx version: nginx/1.12.0
built by gcc 6.3.0 20170415 (Debian 6.3.0-14) 
built with OpenSSL 1.1.0e  16 Feb 2017
TLS SNI support enabled

#2