LNMP | Nginx 1.12/1.13 MySQL | Continuous Deployment library

I run a container in the background using:

Problem: I'm using wordpress on LNMP environment. My site works just fine, but when installing themes / plugins, it does not work and reports a 504 error.

Bottom line:

1. Nginx and php (at least sometimes) communicate fine, since I can see my site display normally.
2. FTP host, user name and password are correct, since when installing themes or plugins, (at least some) files are downloaded into theme or plugin folder (but not the whole theme or plugin).
3. FTP user and Nginx user do have permission to access to the folder, which is owned by FTP user. (FTP & Nginx users are in same group and the folder is set to 775)
4. i tried bigger fastcgi_read_timeout values (e.g. 300), but it did not help.

Details When I run into the problem, I did what people normally do. I went to GOOGLE and set fastcgi_read_timeout to bigger values, as I remember this worked for me on other severs. This, however, only have me wait for 5 min before get the error report (300 s = 5 min). I also tried bigger fastcgi_buffer_size, which also did not work. I think this has something to do with installing theme or plugin, since the web page itself displays fine and error only occurs when installing.When installation failed, I find (not complete plugin) files in plugin folder.

Nginx error log:

I deployed kubernetes 1.5 cluster on the aws, when I want to deploy a LNMP stack on the k8s cluster, but I find that it's difficult to wirte the manifest file, I can't find any examples on k8s official docs. so ,is there any detailed list of manifest parameters and also explan what the parameters meanings?

sorry for my poor english, hope somebody can help me ,thx

I have a LNMP stack with Varnish in front. I have a probe with Varnish and it checks every seconds if the site is running.

It works good but I don't want to log those probes.

Does someone know please how to disable only that log?

Thanks