I´ve tried with @communica configuration, but page doesn´t load (blank page).

This is the log:

2015/01/08 07:50:07 [info] 4050#0: *4 epoll_wait() reported that client prematurely closed connection, so upstream connection is closed too (104: Connection reset by peer) while sending request to upstream, client: 62.14.183.8, server: ec2-54-72-103-90.eu-west-1.compute.amazonaws.com, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php5-fpm.sock:", host: "mysite.com"

Is needed another Nginx or PHP setting?. 

Thanks @communica for share :)

Would be great to accept this pull request to include the nginx reference file in core: https://github.com/Elgg/Elgg/pull/7756

I just wrote my own PR on my own instead of asking again to fix the issues with the proposed PR.

With @Pawel configuration I obtain a blank page too.

This is the log: 

2015/01/09 10:27:37 [info] 1067#0: *24 client 62.14.183.8 closed keepalive connection

What I´m doing wrong?

@Javier I would investigate php deamon logs. Most likely you want to enable logging in php.ini as independent from webserver logs and see what's happening. Might be also issue with php-fpm pool itself. That doesn't strike me as Nginx rewrites issue in your case.

@Pawel I´ve set a debug level logging in php-fpm.conf, but no errors are logged.

I´ve Nginx 1.6.2 and PHP 5.5.9.

Maybe I´ve to update PHP???

I´m going crazy... :(

Error after upgrading from 1.8 to 1.9, images are showing with a ? but CSS works

Failed to load resource: the server responded with a status of 404 (Not Found)

This is related to the nginx.conf but I'm not sure how to fix this.

Never mind, fixed by going here 

site.com/admin/upgrades