server: Raise client_max_body_size for intermittent-tracker

[kkoyung]

WPT Import https://github.com/servo/servo/actions/runs/26348192526 is currently blocked because the intermittent-tracker server rejected queries to https://intermittent-tracker.servo.org/dashboard/query with "HTTP Error 413: Request Entity Too Large" when the request body is too large.

Attempt to unblock it by raising the client_max_body_size setting of nginx for intermittent-tracker servers to 20m. (The default value of global client_max_body_size setting of nginx in NixOS is 10m, instead of 1m in standalone nginx.)

Remark: There was a discussion on similar issue, before migrating CI servers to NixOS. servo/servo#31845

[delan]

thanks for the fix :)

[delan]

deploying:

$ cd server/nixos
$ ./do deploy ci0
$ for i in {1..4}; do ./do deploy ci$i & done

[delan]

putting it in commonHttpConfig does not work:

May 26 06:46:48 ci0 systemd[1]: Starting Nginx Web Server...
May 26 06:46:48 ci0 nginx-pre-start[3726329]: nginx: [emerg] "client_max_body_size" directive is duplicate in /nix/store/01qqxa245b15xp28fxfhvsl5x92wlyqb-nginx.conf:33
May 26 06:46:48 ci0 nginx-pre-start[3726329]: nginx: configuration file /nix/store/01qqxa245b15xp28fxfhvsl5x92wlyqb-nginx.conf test failed
May 26 06:46:48 ci0 systemd[1]: nginx.service: Control process exited, code=exited, status=1/FAILURE
May 26 06:46:48 ci0 systemd[1]: nginx.service: Failed with result 'exit-code'.
May 26 06:46:48 ci0 systemd[1]: Failed to start Nginx Web Server.

[delan]

d2948d4 fixes that. deploying:

$ cd server/nixos
$ ./do deploy ci{0..4}

[delan]

all good, per zulip :)