Install QQ robot on Ubuntu 16.04

Install QQ Robot on Ubuntu#

Looking at it now, it seems that QQ robots only have cool Q robots and Docker can run on Linux.
Then start to install the cool Q robot, the resource consumption is not very large, it takes up about 180M of memory.

Install 酷Q HTTP

First install docker

CoolQ official website provides docker COOLQ Docker 2.0
But I plan to develop some functions through HTTP, so I chose CQHTTP Docker with HTTP plugin made by the big guy
First of all, the pull mirror, I think the download is relatively slow, hang a screen slowly

docker pull richardchien/cqhttp:latest

After the download is complete, start the container for testing

docker run -ti --rm --name coolq \
  - v /home/coolq:/home/user/coolq \  #Mount the host directory to the container for persisting cool Q program files
  - p 9000:9000 \  #noVNC port, used to control CoolQ from the browser
  - p 5700:5700 \  #Port opened by HTTP API plug-in
  - e COOLQ_ACCOUNT=123456 \ #QQ account to log in, optional but recommended
  - e CQHTTP_POST_URL=http://example.com:8080 \  #Incident report address
  - e VNC_PASSWD=111111111 \ #noVnc password
  - e CQHTTP_SERVE_DATA_FILES=yes \  #Allow access to CoolQ data files via HTTP interface
  richardchien/cqhttp:latest

At this point, you can log in to noVNC to view and log in

If there is no problem with the test, just change the --rm in the startup parameters to -d to save and run

docker start coolq
docker stop coolq

supervisord start-up self-start##

Since I have a timed task how often the machine is turned on and off, but coolq will not start automatically after booting, select supervisord to start automatically
Pay attention to autorestart=false, because after starting coolq, it will be exit(0), so supervisord will keep trying to restart, we only need him to start it once.

[ program:coolq]
command=docker start coolq
stderr_logfile=/var/log/supervisor/error_coolq.log
stdout_logfile=/var/log/supervisor/coolq.log
directory=/home
autostart=true
user=docker
autorestart=false

Configuration information##

Account configuration file###

Note that the report address cannot be 127.0.0.1, because the container has its own 127.0.0.1, in general, the machine ip can be used as the report address

[123456789]
use_ws =false
use_ws_reverse =false
serve_data_files = yes
access_token =11111111111111111
log_level = error
post_url = http://example.com:8080
secret =11111111111111111
show_log_console =false
event_filter = filter.json

Filter and report HTTP configuration files###

• Private message
• Non-anonymous group messages of @ Robot
• Non-anonymous'coming out' group messages
• Add friend message, add group invitation, add group request

{". or":[{"message_type":"private"},{"message_type":"group","anonymous":{".eq":null},"raw_message":{".regex":"CQ:at,qq=2450184313|^come out$"}},{"post_type":"request"}]}

Firewall##

Because Docker directly changes iptables and has a higher priority than ufw, you cannot control the opening and closing of its listening ports through ufw, you need to manually configure Docker information

# vim /etc/default/ufw
DEFAULT_FORWARD_POLICY="ACCEPT"

# vim /etc/ufw/before.rules
# * Add the following content before the filter
# 172.17.0.0 /16 is the docker bridge address, which may be different
* nat
: POSTROUTING ACCEPT [0:0]-A POSTROUTING !-o docker0 -s 172.17.0.0/16-j MASQUERADE
COMMIT

# vim /etc/default/docker
# Add this configuration information
DOCKER_OPTS="--dns 8.8.8.8 --dns 8.8.4.4 --iptables=false"

# vim /etc/docker/daemon.json
# Create without this file
{" iptables":false}

Restart daemon, docker, ufw

systemctl daemon-reload && systemctl restart docker && ufw reload