Python – uwsgi broken pipe – django, nginx

djangonginxpythontastypieuwsgi

I'm randomly (and consistently) getting a broken pipe in uwsgi… shown below. Any idea what could be causing this or how I can debug?

I'm on django (tastypie), uwsgi, nginx, and am running an m3.medium on aws (ubuntu 14.04).

[pid: 1516|app: 0|req: 548/1149] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:11 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 20 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1517|app: 0|req: 594/1150] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:12 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 15 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1516|app: 0|req: 549/1151] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:13 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 15 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1516|app: 0|req: 550/1152] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:13 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 14 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1517|app: 0|req: 595/1153] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:14 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 15 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1516|app: 0|req: 551/1154] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:14 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 14 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1517|app: 0|req: 596/1155] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:15 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 12 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1516|app: 0|req: 552/1156] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:15 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 12 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
Wed Jun 18 16:11:17 2014 - uwsgi_response_writev_headers_and_body_do(): Broken pipe [core/writer.c line 287] during GET /api/v1/clock/?format=json (10.0.0.204)
IOError: write error
[pid: 1512|app: 0|req: 1/1157] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:16 2014] GET /api/v1/clock/?format=json => generated 0 bytes in 1460 msecs (HTTP/1.1 200) 4 headers in 0 bytes (0 switches on core 0)
announcing my loyalty to the Emperor...
Wed Jun 18 20:11:17 2014 - [emperor] vassal api.ini is now loyal
[pid: 1516|app: 0|req: 553/1158] 10.0.0.159 () {42 vars in 1039 bytes} [Wed Jun 18 16:11:33 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 14 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1516|app: 0|req: 554/1159] 10.0.0.204 () {46 vars in 908 bytes} [Wed Jun 18 16:11:41 2014] GET /api/v1/clock/ => generated 1298 bytes in 14 msecs (HTTP/1.0 200) 4 headers in 119 bytes (1 switches on core 0)

i'm noticing the number on whatever is related drop to a very low number too. Note the second request here — 2/1303. This request timed out.

[pid: 1516|app: 0|req: 624/1302] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:41:09 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 12 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1512|app: 0|req: 2/1303] 10.0.0.204 () {42 vars in 1039 bytes} [Wed Jun 18 16:41:10 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 50 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1516|app: 0|req: 625/1304] 10.0.0.159 () {42 vars in 1039 bytes} [Wed Jun 18 16:41:29 2014] GET /api/v1/clock/?format=json => generated 1298 bytes in 17 msecs (HTTP/1.1 200) 4 headers in 119 bytes (1 switches on core 0)
[pid: 1517|app: 0|req: 668/1305] 10.0.0.204 () {46 vars in 908 bytes} [Wed Jun 18 16:41:31 2014] GET /api/v1/clock/ => generated 1298 bytes in 18 msecs (HTTP/1.0 200) 4 headers in 119 bytes (1 switches on core 0)

UPDATED: nginx.conf

user www-data;
worker_processes 1;
pid /run/nginx.pid;

events {
    worker_connections 1024;
    # multi_accept on;
}

http {

    client_body_timeout 12;
    client_header_timeout 12;
    keepalive_timeout 15;
    send_timeout 10;
    client_max_body_size 8m;

    ##
    # Basic Settings
    ##

    sendfile on;
    tcp_nopush on;
    tcp_nodelay on;
    types_hash_max_size 2048;
    # server_tokens off;

    # server_names_hash_bucket_size 64;
    # server_name_in_redirect off;

    include /etc/nginx/mime.types;
    default_type application/octet-stream;

    ##
    # Logging Settings
    ##

    #access_log off;
    access_log /var/log/nginx/access.log;
    error_log /var/log/nginx/error.log;

    ##
    # Gzip Settings
    ##

    gzip on;
    gzip_disable "msie6";

    # gzip_vary on;
    # gzip_proxied any;
    # gzip_comp_level 6;
    # gzip_buffers 16 8k;
    # gzip_http_version 1.1;
    # gzip_types text/plain text/css application/json application/x-javascript text/xml application/xml application/xml+rss text/javascript;

    ##
    # nginx-naxsi config
    ##
    # Uncomment it if you installed nginx-naxsi
    ##

    #include /etc/nginx/naxsi_core.rules;

    ##
    # nginx-passenger config
    ##
    # Uncomment it if you installed nginx-passenger
    ##

    #passenger_root /usr;
    #passenger_ruby /usr/bin/ruby;

    ##
    # Virtual Host Configs
    ##

    include /etc/nginx/conf.d/*.conf;
    include /etc/nginx/sites-enabled/*;
}

this specific site's virtual config

upstream django {
    server unix:/tmp/domain.sock;
}

server {
    listen 80;

    server_name domain.com;

    return 301 https://$host$request_uri;
}

server {
    listen 443;
    server_name domain.com;

    location /static {
        alias /home/ubuntu/domain/static;
    }

    location / {
        proxy_set_header X-Forwarded-Proto https;

        uwsgi_pass django;
        include /etc/nginx/uwsgi_params;
    }
}

uwsgi config (vassals)

[uwsgi]
env             = DEBUG=False
env             = DB_ENVIRONMENT=production
env             = NEW_RELIC_CONFIG_FILE=config/newrelic.ini
env             = NEW_RELIC_ENVIRONMENT=production
chdir           = /home/ubuntu/domain
home            = /home/ubuntu/domain/venv
module          = domain.wsgi
processes       = 20
uid             = www-data
gid             = www-data
chmod-socket    = 666
socket          = /tmp/domain.sock
stats           = /tmp/domain.stats.sock

contend of /etc/rc.local which kick off the uwsgi process at boot

#!/bin/sh -e
/usr/local/bin/uwsgi --emperor /etc/uwsgi/vassals --logto /var/log/uwsgi/emperor.log
exit 0

Best Answer

you can safely ignore them, they are triggered by the client (or nginx) disconnecting in the middle of the request. As the response times are really low, it is very probably a client disconnection. Btw, post your nginx and uWSGI configuration for being sure.

Related Topic