0

I have a Kirby CMS website (PHP server behind nginx reverse proxy) running in Amazon EFS. The most recent update of the site lead to 502 Bad Gateway and 504 Gateway Timeout errors.

Looking at the nginx logs, I think that 502s are caused by this:

2022/08/02 08:41:01 [error] 2887#2887: *11071 connect() to unix:/run/php-fpm/www.sock failed (11: Resource temporarily unavailable) while connecting to upstream, client: 172.31.25.144, server: , request: "...", upstream: "fastcgi://unix:/run/php-fpm/www.sock:", host: "..."

...while 504s by this:

2022/08/02 08:41:20 [error] 2887#2887: *10898 upstream timed out (110: Connection timed out) while reading response header from upstream, client: 172.31.25.144, server: , request: "...", upstream: "fastcgi://unix:/run/php-fpm/www.sock", host: "..."

Looking at the EFS dashboard in my sandbox environment, I can see a surge of writes to the filesystem during my load test period:

Amazon EFS graphs

I don't understand where those writes come from.

Question: Is there a way to have an access.log file, similar to that of nginx? I want to see reads, writes, and metadata requests to the NFS, so I can figure out what's happening.

I came across nfslogd, but it appears to be software for Oracle Solaris. Is there an equivalent for Amazon Linux 2/3.3.15?

1
  • a) do you host /run/php-fpm/www.sock on nfs, (b) use tcpdump + wireshark
    – kofemann
    Aug 31 at 15:43

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy