Trying to use this image in a docker-compose setup with cassandra is impossible. The issue is that cassandra takes so long to startup and kong bails if cassandra isn't running.

I solved this by adding monit to the base image https://hub.docker.com/r/articulate/kong-monit/ and https://github.com/articulate/docker-kong-monit

We could easily roll this back into this repo by creating a secondary Dockerfile.monit and then having docker hub build that along with the other Dockerfile. The normal one will be tagged how it is right now, but we can have the monit version tagged as monit-latest and monit-0.5.3.

This gives the person using the image the choice.

I'd be happy to PR the changes and work with whoever needed to get this done.

Hello,

I'm trying to configure hmac-auth and wrote some script :

#!/usr/bin/perl

use strict;
use warnings;

use HTTP::Date;
use MIME::Base64;
use Digest::SHA qw(sha1 sha1_hex sha1_base64 hmac_sha1_base64);

my $date = time2str();
my $key = 'hmactest';

#my $hmac = Digest::HMAC_SHA1->new($key);
#$hmac->add($date);
#my $digest = $hmac->digest;

my $hmac = hmac_sha1_base64("date: $date\n",$key);
system( "curl --url http://localhost:8000/ --header 'date: $date' --header 'Host: example.com' --header 'Authorization:  hmac usename=\"user1\", algorithm=\"hmac-sha1\", headers=\"X-Date \", signature=\"$hmac\"'\n"");

but got error :
{"message":"HMAC signature cannot be verified"}

What I'm doing wrong ?

Hi All,

I am trying to get Kong properly working on Tutum/Docker Cloud (as it is now called).

https://cloud.docker.com

The issue is that kong always fails when using docker links on Tutum/DockerCloud.

It appears that DNSmasq is not using the search element of the resolve.conf to add the FQDN to the request and as such it just fails.

So if I set in kong.yml :

contact_points:
    - "cassandra:9042"

Kong will fail to start unable to connect to cassandra. It should resolve to casasndra....local.dockerapp.io

Example of resolve.conf on docker cloud:

search .<container-id>..local.dockerapp.io
nameserver 8.8.8.8
nameserver 8.8.4.4

This issue also prevents you using microservices behing kong via docker links as they do not resolve the same way as cassandra does not.

E.g http://chatservice should resolve to http://chatservice..local.dockerapp.io

container-id changes each time you redeploy or bring new containers up so you can not hard code this, the DNS for . is only active for as long as the container is.

How can I get kong/dnsmasq to properly use the search option in resolve.conf?

If we can solve this issue then I will no longer have to host a custom version of Kong for Docker Cloud

https://github.com/holidaywatchdog/docker-cloud-kong

I should mention that cassandra resolves correctly from within the docker container just not via dnsmasq

root@api-gateway-1 /]# ping cassandra
PING cassandra.a72c1fd7-97e3-423b-ac69-d02a5dee5d8e.local.dockerapp.io (10.7.0.2
2) 56(84) bytes of data.
64 bytes from 10.7.0.22: icmp_seq=1 ttl=64 time=1.79 ms

The only short term work around I have for this is to use a dummy container in the stack and link it to all the services you need to use with kong. Then use that containers FQDN to reference the services. As long as this container never goes down the DNS should work.

When trying to link kong to cassandra db via compose up, Kong is not able to connect to cassandra container. However same is working perfectly when i try to link manually from command line.
I also tried with docker-machine ip in kong.yml, but got same error.

Error after compose up:
"
Tushars-MBP:docker-kong tushar$ docker logs 534e5cc8b4a0
[INFO] kong 0.8.2
[INFO] Using configuration: /etc/kong/kong.yml
[INFO] Setting working directory to /usr/local/kong
[INFO] database...........cassandra contact_points=kong-database:9042 data_centers= ssl=verify=false enabled=false port=9042 timeout=5000 replication_strategy=SimpleStrategy keyspace=kong replication_factor=1 consistency=ONE
[INFO] Leaving cluster..
[ERR] All hosts tried for query failed. kong-database:9042: connection refused for socket with peer kong-database:9042.
[ERR] Could not start Kong
"
Compose file:
"
kong-database:
image: cassandra:2.2
hostname: kong-database
ports:

• "9042:9042"

kong:
image: mashape/kong:0.8.2
ports:

• "8000:8000"
• "8001:8001"
• "8443:8443"

links:

• kong-database:kong-database

environment:

• DATABASE=cassandra
  "

Docker info:

"
Tushars-MBP:docker-kong tushar$ docker info
Containers: 0
Running: 0
Paused: 0
Stopped: 0
Images: 44
Server Version: 1.11.2
Storage Driver: aufs
Root Dir: /mnt/sda1/var/lib/docker/aufs
Backing Filesystem: extfs
Dirs: 97
Dirperm1 Supported: true
Logging Driver: json-file
Cgroup Driver: cgroupfs
Plugins:
Volume: local
Network: bridge null host
Kernel Version: 4.4.12-boot2docker
Operating System: Boot2Docker 1.11.2 (TCL 7.1); HEAD : a6645c3 - Wed Jun 1 22:59:51 UTC 2016
OSType: linux
Architecture: x86_64
CPUs: 1
Total Memory: 995.9 MiB
Name: handler
ID: BJHX:OKR3:YWIU:MAX5:PIR4:2I2J:C36B:TEHT:FQEJ:6IYT:6THT:EJHN
Docker Root Dir: /mnt/sda1/var/lib/docker
Debug mode (client): false
Debug mode (server): true
File Descriptors: 12
Goroutines: 30
System Time: 2016-08-11T05:57:34.512054505Z
EventsListeners: 0
Username: tushardwivedi
Registry: https://index.docker.io/v1/
Labels:
provider=virtualbox
"

Currently, nginx/openresty is running as root. This is not considered good riddance. Instead, the user running Kong should be a non-privileged user, perhaps called "kong".

Kong 0.8.0 has just released. Will it be available as docker image?

Have the instructions in the README been run against docker 1.12? (doing this on a mac)

Both with and without an overlay net I can't seem to get kong to connect to cassandra

docker run -d --name kong-database \
                -p 9042:9042 \
                cassandra:2.2

docker run -d --name kong \
    -e "DATABASE=cassandra" \
    --link kong-database:kong-database \
    -p 8000:8000 \
    -p 8443:8443 \
    -p 8001:8001 \
    -p 7946:7946 \
    -p 7946:7946/udp \
    --security-opt seccomp:unconfined \
    mashape/kong

[INFO] kong 0.8.3
[INFO] Using configuration: /etc/kong/kong.yml
[INFO] Setting working directory to /usr/local/kong
[INFO] database...........cassandra contact_points=kong-database:9042 data_centers= ssl=verify=false enabled=false port=9042 timeout=5000 replication_strategy=SimpleStrategy keyspace=kong replication_factor=1 consistency=ONE
[INFO] Leaving cluster..
[ERR] All hosts tried for query failed. kong-database:9042: connection refused for socket with peer kong-database:9042.
[ERR] Could not start Kong

It looks like 0.9.1 and 0.9.2 should have been pushed to docker hub, but I'm not seeing them up there.

Seems that running kong and cassandra in docker containers eat disk space until there is no space left on my virtual machine.

Running df -h on the kong container: docker exec kong df -h

Filesystem      Size  Used Avail Use% Mounted on
none             48G   16G   29G  36% /
tmpfs           2.0G     0  2.0G   0% /dev
tmpfs           2.0G     0  2.0G   0% /sys/fs/cgroup
/dev/sda1        48G   16G   29G  36% /etc/kong
shm              64M     0   64M   0% /dev/shm

At some point /dev/sda1 will go up 100% use. Cassandra/Kong logs does not show anything particular. Any idea what might be wrong?

Using cassandra:2.2.5 image and mashape/kong image. Also docker info shows:

Containers: 15
 Running: 2
 Paused: 0
 Stopped: 13
Images: 55
Server Version: 1.10.1
Storage Driver: aufs
 Root Dir: /mnt/sda1/var/lib/docker/aufs
 Backing Filesystem: extfs
 Dirs: 232
 Dirperm1 Supported: true
Execution Driver: native-0.2
Logging Driver: json-file
Plugins:
 Volume: local
 Network: bridge null host
Kernel Version: 4.1.17-boot2docker
Operating System: Boot2Docker 1.10.1 (TCL 6.4.1); master : b03e158 - Thu Feb 11 22:34:01 UTC 2016
OSType: linux
Architecture: x86_64
CPUs: 1
Total Memory: 3.859 GiB
Name: dinghy
ID: IOWU:RZYU:5UA3:FX4Z:K7TZ:APJV:STOA:PKBT:DFAY:575B:SEJT:JR2S
Debug mode (server): true
 File Descriptors: 36
 Goroutines: 61
 System Time: 2016-02-17T14:22:12.045636927Z
 EventsListeners: 1
 Init SHA1:
 Init Path: /usr/local/bin/docker
 Docker Root Dir: /mnt/sda1/var/lib/docker
Username: gottfrois
Registry: https://index.docker.io/v1/
Labels:
 provider=virtualbox

Note: Running only cassandra image does not go up 100% disk use. It seems that it's kong that eat the disk space.

It appears "cassandra" is hard-wired as the hostname for the cassandra database. Using --link, the IP and port is available to the container - but it appears the config file does not expand environment variables? It would be nice if --link was sufficient without having to use a hardcoded name and export the cassandra port explicitly.

I just did docker pull kong after having some issues getting Kong to install via docker-compose up, now I'm getting...

Using default tag: latest
Pulling repository docker.io/library/kong
Error: image library/kong not found

Hello,

I'm using dockerized kong and start it well. Inside container all of requests works good, but when I send request outside container ( like curl -v --url http://m.y.i.p:8000/ --header 'Host: example.com' ) kong said :

<cross-domain-policy>
<allow-access-from domain="*" to-ports="*"/>
</cross-domain-policy>

How can I open my api requests to outside ?

I moved this comment from another issue to its own issue to hope to get a response/discussion.

We currently cannot move to Kong 0.9.0 because the new Docker does not run nginx as root user (which is good), but user nobody seems not allowed to write to /dev/stdout.
In 0.8.3 we used the file-log plugin to log the output (instead of the method described by Jalberto in #40), since this suggestion does not seem to work.

This now gives the following error on each request in the nginx error log within the docker:
2016/08/28 16:01:18 [error] 76#0: [lua] handler.lua:56: [file-log] failed to open the file: Permission denied, context: ngx.timer, client: xxx.xxx.xxx.xxx, server: 0.0.0.0:8000

Does anyone have another solution?

I have checked out your code and tried starting with docker-compose up. But it fails saying
ERROR: for nginx-lb Cannot start service nginx-lb: driver failed programming external connectivity on endpoint compose_nginx-lb_1 (45fe13a6dd43c580f6bf946ee14ba0630cfca2ae3c73ce23e3b712a5a1ac5a75): Bind for 0.0.0.0:8443 failed: port is already allocated

not a docker-kong bug/issue but starting the cassandra container errors out for me

vagrant@vagrant:kong|next⚡ ⇒  docker ps -a
CONTAINER ID        IMAGE               COMMAND             CREATED             STATUS              PORTS               NAMES
vagrant@vagrant:kong|next⚡ ⇒  docker images
REPOSITORY          TAG                 IMAGE ID            CREATED             SIZE
php                 5-apache            a13e60b97983        8 weeks ago         483.8 MB
unirest-issue99     latest              a13e60b97983        8 weeks ago         483.8 MB
vagrant@vagrant:kong|next⚡ ⇒  docker run -d --name kong-database \
>                 -p 9042:9042 \
>                 cassandra:2.2
Unable to find image 'cassandra:2.2' locally
2.2: Pulling from library/cassandra

51f5c6a04d83: Pull complete 
a3ed95caeb02: Pull complete 
2a21294c613b: Pull complete 
4d679f1ad019: Pull complete 
3b49a7450c9a: Pull complete 
1e88c3a1e691: Pull complete 
b464a0d9b04f: Pull complete 
8b59c8e0cfa3: Pull complete 
bdbda7d0c45f: Pull complete 
f0664e7f1560: Pull complete 
01554e7bf5c4: Pull complete 
Digest: sha256:c33aeb965f108ac458d1fe8c9c7e563c1cf62cbdee16ed7ff2c658e24a0fb53f
Status: Downloaded newer image for cassandra:2.2
684b7a8e819ed376c8429d89810ec740a044bf9d07740834fe19d51b36cfb01d
vagrant@vagrant:kong|next⚡ ⇒  docker ps -a
CONTAINER ID        IMAGE               COMMAND                  CREATED             STATUS                      PORTS               NAMES
684b7a8e819e        cassandra:2.2       "/docker-entrypoint.s"   34 seconds ago      Exited (1) 33 seconds ago                       kong-database
vagrant@vagrant:kong|next⚡ ⇒  docker logs kong-database
OpenJDK 64-Bit Server VM warning: INFO: os::commit_memory(0x00000000a6600000, 1329594368, 0) failed; error='Cannot allocate memory' (errno=12)
#
# There is insufficient memory for the Java Runtime Environment to continue.
# Native memory allocation (malloc) failed to allocate 1329594368 bytes for committing reserved memory.
# An error report file with more information is saved as:
# /tmp/hs_err_pid1.log

Hi guys,

When following the instructions for getting kong up and running on docker from scratch (the one at https://getkong.org/install/docker/), you run into the following little thing when on Mac OS X or Windows, which actually use VirtualBox for running the default image as docker host: The curl on localhost will not work, as the actual docker host is not localhost, but the docker host inside VirtualBox.

Instead, you have to issue your curl command against the IP your $DOCKER_HOST connection points to. Then it'll work. With my installation on Windows, that's

curl http://192.168.99.100:8001

This might not be an issue for anyone beyond n00b status, but I stumbled over it and had to give it some extra thought...

Best regards,
Martin

Hello,

I've created a docker-compose setup, together with a 'install kong from git source' Dockerfile.
https://github.com/vikingco/docker-kong

Is that something you guys can use? Then we might consider integrating my approach into this repository (and deprecating my repository)

Cheers

I miss an example for adding an api running on the docker host.
It took me quite a long time to figure out that your upstream_url has to be pointing to the docker bridge ip of the docker host (if running docker in bridge mode)
I know that this might not be new knowledge to people with docker experience, but arriving at docker for running kong i missed the following example:

maybe an example could be python based:
if kong is running in docker and the docker host has ip 172.17.42.1 on the docker bridge interface.

(get docker host bridge ip with: ifconfig docker0 | grep 'inet addr:' | cut -d: -f2 | awk '{ print $1}' )

make simple html index:

cd /tmp/
echo "hello world" > index.html

start api endpoint server:

sudo python -m SimpleHTTPServer 9876

add it to kong:

curl -i -XPOST "http://127.0.0.1:8001/apis/" --data 'name=example' --data     
'upstream_url=http://172.17.42.1:9876/' --data 'request_host=example.com' 

test it out

curl -XGET --header 'Host: example.com' "http://localhost:8000/"

and it prints:
"hello world"

I've run 0.8.0 with custom config to postgres as follows:

$ docker run -d \
    -v /path/to/your/kong/configuration/directory/:/etc/kong/ \
    -p 8000:8000 \
    -p 8443:8443 \
    -p 8001:8001 \
    -p 7946:7946 \
    -p 7946:7946/udp \
    --security-opt seccomp:unconfined \
    --name kong \
    mashape/kong

kong.yml

....
database: "postgres"
....

However, setup.sh appends the DATABASE variable to config when run.

kong.yml

....
database: "postgres"
....
database: "cassandra"

It doesn't pickup my postgres in kong.yml unless the DATABASE environment variable is set to "postgres" as well (it's set to cassandra by default). Even then, database: "postgres" is appended to my kong.yml every time I restart the container.

Kong/kong#400

As this image by default doesn't throw logs to std* is harder to debug problems.

Nginx official docker use this trick to enalbe docker-logs:

# forward request and error logs to docker log collector
RUN ln -sf /dev/stdout /var/log/nginx/access.log \
    && ln -sf /dev/stderr /var/log/nginx/error.log

I have a custom Kong configuration and Cassandra instance running. Kong succesfully connects to my database, and logs [OK] Started, but then immediately exits with status 0. This happens when the container is started from Marathon, or manually. My logs and configuration can be found here, scroll to the bottom for the manual run.

I know the Cassandra database to be good, I can manually connect to it using cqlsh.

Any pointers? Any ideas why this might happen? Or how I can debug?

I have successfully started up things using the docker-compose file can send requests to kong.

I have configure the hmac plugin, but all my requests are 403-ing.

Have done some investigation the system time in the docker container is set incorrectly.

The container time is Tue Aug 23 15:48:20 UTC 2016 but my system time is Wed 24 Aug 2016 23:43:31 BST

Hi there,

I'm having a recurrent problem on my kong 0.7.0 containers that a few seconds after the container goes up, it crashes with the message "Can't find nginx"

Here is the full log messages on the container startup

�[0m�[34m[INFO] �[0m�[0mKong 0.7.0
�[0m�[34m[INFO] �[0m�[0mUsing configuration: /etc/kong/kong.yml
�[0m�[34m[INFO] �[0m�[0mSetting working directory to /usr/local/kong
�[0m�[34m[INFO] �[0m�[0mdatabase...........cassandra auth=password= username= timeout=5000 data_centers= password= ssl=verify=false enabled=false contact_points=: replication_strategy=SimpleStrategy username= replication_factor=1 keyspace=kong
�[0m�[34m[INFO] �[0m�[0mdnsmasq............address=127.0.0.1:8053 dnsmasq=true port=8053
�[0m�[34m[INFO] �[0m�[0mserf ..............-profile=wan -rpc-addr=127.0.0.1:7373 -event-handler=member-join,member-leave,member-failed,member-update,member-reap,user:kong=/usr/local/kong/serf_event.sh -bind=0.0.0.0:7946 -advertise=<my_ip>:7946 -node=_0.0.0.0:7946 -log-level=err
�[0m�[34m[INFO] �[0m�[0mTrying to auto-join Kong nodes, please wait..
�[0m�[33m[WARN] �[0m�[0mCannot join <cassandra_node_ip>:. If the node does not exist anymore it will be automatically purged.
�[0m�[34m[INFO] �[0m�[0mSuccessfully auto-joined <another_cassandra_node_ip>:
�[0m�[34m[INFO] �[0m�[0mAuto-generating the default SSL certificate and key...
�[0m�[34m[INFO] �[0m�[0mLeaving cluster..
�[0m�[31m[ERR] �[0m�[0mCan't find nginx
�[0m�[31m[ERR] �[0m�[0mCould not start Kong

Can someone help me please?

I wanted to set upstream based on HTTP method for same URLs, for example I have /posts URL and wanted to send request to different hosts when the method is GET,POST,DELETE or something else. As I read the docs there are options based on URL or host.
Is it any possibility to have this kind of routes?

Hi there,

I'm looking around on how to create custom-plugin, but as far as I understood, I'll have to build the kong source to include my custom plugin.
Since this container always download the rpm file, I cannot compile my plugin's source.

I'm following the steps on this tutorial: http://streamdata.io/blog/developing-an-helloworld-kong-plugin/

If I misunderstood something, please help me.

Thanks!
Francisco Andrade

Would be great to have a tag for ARM so raspberryPI adn scaleway C1 can be used as kong hosts

Every time I try to start Kong, I get the following error...

[INFO] kong 0.8.1
[INFO] Using configuration: /etc/kong/kong.yml
[INFO] Setting working directory to /usr/local/kong
[INFO] database...........cassandra contact_points=kong-database:9042 data_centers= ssl=verify=false enabled=false port=9042 timeout=5000 replication_strategy=SimpleStrategy keyspace=kong replication_factor=1 consistency=ONE
[INFO] Leaving cluster..
[ERR] All hosts tried for query failed. kong-database:9042: connection refused for socket with peer kong-database:9042.
[ERR] Could not start Kong

A docker ps shows the following

a831fdfb1751 cassandra:2.2 "/docker-entrypoint.s" 53 minutes ago Up 53 minutes 7000-7001/tcp, 7199/tcp, 9160/tcp, 0.0.0.0:9042->9042/tcp kong-database

I've already got Cassandra running as instructed with a link named kong-database. I can shell into kong-database and run cqlsh fine.

Anyone have any idea what I can try to get closer to uncovering the problem?

I'm trying to start a Kong cluster as Docker containers, but I can't manage to create the cluster because the cluster_listen_address is a Docker IP (172.x.x.x) that don't communicate with a container running in another host.
I've managed to sucessfully configure the cluster when using the host network in Docker, but my goal is to use docker network and the mapped ports.
When I do that, use customized ports and set the cluster_listen_address to be the Host IP, and not a Docker IP, I get this error:

[INFO] Kong 0.7.0
[INFO] Using configuration: /etc/kong/kong.yml
[INFO] Setting working directory to /usr/local/kong
[INFO] database...........cassandra auth=password=cassandra username=cassandra timeout=5000 data_centers= password=cassandra ssl=verify=false enabled=false contact_points=10.166.69.149:39044 replication_strategy=SimpleStrategy username=cassandra replication_factor=1 keyspace=kong
[INFO] dnsmasq............address=127.0.0.1:8053 dnsmasq=true port=8053
[INFO] Leaving cluster..
[ERR] Could not start serf: Failed to start the Serf agent: Error creating Serf: Failed to create memberlist: Failed to start TCP listener. Err: listen tcp 10.166.69.149:37947: bind: cannot assign requested address
[ERR] Could not start Kong

When I use the --net=host option with docker, the "10.166.69.149:37947" cluster_listen_address successfully appears in my Kong.nodes table.

I'm having issues with the compose version, because the container is always restarting... How do I know if the problem is my machine or the image? Is there any command that can help me printing the error?

Thank you for your time.

PS: I used your compose version without making changes into it. So That is why I'm asking through this platform.

Hello,

I hope I just missed something but I have a problem running KONG with docker. If I tries to run it without -t (tty) it doesn't output any logs.

If I run it with the -t (tty) it get stuck in this step: [INFO] Auto-generating the default SSL certificate and key...

I am trying to run integration tests against KONG in a container. And the same thing there, it hangs as soon as a child process is spawn.

Hi there,

I have the following set-up:

cassandra:
  image: cassandra:2.2.5

kong:
  image: mashape/kong
  links:
    - cassandra
  security_opt:
    - seccomp:unconfined
  ports:
    - "8000:8000"
    - "8443:8443"
    - "8001:8001"
    - "7946:7946"
    - "7946:7946/udp"

core-api:
  build: ./applications/core/api
  ports:
    - "8080:8080"
  links:
    - core-datastore 
    - core-cache
    - elasticsearch
  environment:
    - DB_NAME=bandzest
    - VIRTUAL_HOST=core.bandzest.dev
  restart: always

Then when I run curl -i -XGET http://bandzest.dev:8000 --header 'Host: core.bandzest.dev' -d 'request_path=api/v1/artists'

I get

HTTP/1.1 502 Bad Gateway
Date: Thu, 31 Mar 2016 00:20:13 GMT
Content-Type: text/plain; charset=UTF-8
Transfer-Encoding: chunked
Connection: keep-alive
Server: kong/0.7.0

An invalid response was received from the upstream server

Whereas if I go straight to the API, I get a proper response.

Any pointers? Thanks in advance.

When I test whether the container is running normally, I met the following error:

{"message":"Cassandra error: Error during UNIQUE check: Cassandra error: cassandra could not be resolved (2: Server failure)"}

The following is the details:

[root@9fdc4a788c41 opt]# curl -i -X POST
-data 'name=mockbin'
--data 'target_url=http://mockbin.com/'
--data 'public_dns=mockbin.com'> --url http://localhost:8001/apis/ \

--data 'name=mockbin'
--data 'target_url=http://mockbin.com/' \

HTTP/1.1 400 Bad Request
Date: Wed, 29 Jul 2015 06:43:26 GMT
Content-Type: application/json; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive
Server: kong/0.2.1

{"public_dns":"public_dns is required"}
[root@9fdc4a788c41 opt]# curl -i -X POST
-data 'name=mockbin'
--data 'target_url=http://mockbin.com/'
--data 'public_dns=mockbin.com'> --url http://localhost:8001/apis/ \

--data 'name=mockbin'
--data 'target_url=http://mockbin.com/'
--data 'public_dns=mockbin.com'
HTTP/1.1 500 Internal Server Error
Date: Wed, 29 Jul 2015 06:44:32 GMT
Content-Type: application/json; charset=utf-8
Transfer-Encoding: chunked
Connection: keep-alive
Server: kong/0.2.1

{"message":"Cassandra error: Error during UNIQUE check: Cassandra error: cassandra could not be resolved (2: Server failure)"}

who can help me to go through it pls?

hi. would it be possible to have some support about configuring docker's kong image?
i need to add a specific lua script that should be runned by openresty (something i want to run after kong.access() call), so i updated nginx-kong.conf and add my command afteraccess_by_lua_block, i updated my own Dockerfile like this.

COPY config/nginx-kong.conf /usr/local/kong/

but it seems the file is overwritten.
is it the way to go? if yes what is the appropriate file to overwrite?
or should i use templates or something else?
thank you

Most of Docker Official Images are based on Debian or Alpine while this Kong image is based on CentOS.
Is there any specific reason not to use Debian?

Kong has just released version 0.9.6 but I coudn't see that tag on Docker Hub

I'm running the 0.6.0 docker container, and getting this output:

...
kong_1       | [INFO] Migrating response-ratelimiting on keyspace "kong" (cassandra)
kong_1       | [INFO] response-ratelimiting migrated up to: 2015-08-21_init_response-rate-limiting
kong_1       | [INFO] Migrating hmac-auth on keyspace "kong" (cassandra)
kong_1       | [INFO] hmac-auth migrated up to: 2015-09-16-132400_init_hmacauth
kong_1       | [INFO] dnsmasq............address=127.0.0.1:8053 dnsmasq=true port=8053
kong_1       | [INFO] Auto-generating the default SSL certificate and key...
kong_1       | [INFO] nginx .............admin_api_listen=0.0.0.0:8101 proxy_listen=0.0.0.0:8101 proxy_listen_ssl=0.0.0.0:8543
kong_1       | [INFO] serf ..............-profile=wan -rpc-addr=127.0.0.1:7373 -event-handler=member-join,member-leave,member-failed,member-update,member-reap,user:kong=/usr/local/kong/serf_event.sh -bind=0.0.0.0:7946 -node=1e2396c98d96_0.0.0.0:7946 -log-level=err
kong_1       | [INFO] Trying to auto-join Kong nodes, please wait..
kong_1       | [WARN] Cannot auto-join the cluster because no nodes were found
kong_1       | [OK] Started
devenv_kong_1 exited with code 0

I'm aware the 0.6.0 kong forces us to now use serf, but how does it relate to docker? Do I need a serf daemon on the kong container? Or a separate container with serf? Or how do I disable serf for development/testing purposes?

Im trying to run the example provided with Consul and Nginx using the compose. Once everything was up and running I tried to access Kong from the Kong Dashboard through Nginx instance. What would the URL to access Kong admin API? I tried http://nginx-lb:8001 and its not accessible. Im using Kong 0.9. I also tried to attach to the nginx container and ping'ed the "proxy_pass" (kong-8001) but got a host not found error. Any idea what im doing wrong? What would be the URL to access the kong admin api through the nginx reverse proxy?

I'm trying to get docker compose working but it always brings up the kong container before the postres one... Apparently it's bad design for microservices to die if their dependencies are not up, they should keep trying to connect. Can you make Kong behave like this please?

In any event, this is all making it impossible for me to make this docker-compose work. Can you please fix/provide workaround/comment.

thanks

Hello,

I'm currently investigating how we can run docker-kong in a docker-compose setup (https://docs.docker.com/compose/)

My current docker-compose.yml:

cassandra:
  image: mashape/cassandra
  ports:
    - "9042:9042"

kong:
  image: mashape/kong
  links:
    - cassandra
  ports:
    - "8001:8001"
    - "8000:8000"

Unfortunatly, this doesn't work:

$ docker-compose up
Creating dockerkong_cassandra_1...
Creating dockerkong_kong_1...
Attaching to dockerkong_cassandra_1, dockerkong_kong_1
cassandra_1 | CompilerOracle: inline org/apache/cassandra/db/AbstractNativeCell.compareTo (Lorg/apache/cassandra/db/composites/Composite;)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/db/composites/AbstractSimpleCellNameType.compareUnsigned (Lorg/apache/cassandra/db/composites/Composite;Lorg/apache/cassandra/db/composites/Composite;)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/utils/ByteBufferUtil.compare (Ljava/nio/ByteBuffer;[B)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/utils/ByteBufferUtil.compare ([BLjava/nio/ByteBuffer;)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/utils/ByteBufferUtil.compareUnsigned (Ljava/nio/ByteBuffer;Ljava/nio/ByteBuffer;)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/utils/FastByteOperations$UnsafeOperations.compareTo (Ljava/lang/Object;JILjava/lang/Object;JI)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/utils/FastByteOperations$UnsafeOperations.compareTo (Ljava/lang/Object;JILjava/nio/ByteBuffer;)I
cassandra_1 | CompilerOracle: inline org/apache/cassandra/utils/FastByteOperations$UnsafeOperations.compareTo (Ljava/nio/ByteBuffer;Ljava/nio/ByteBuffer;)I
kong_1      | [INFO] Using configuration: /etc/kong/kong.yml
kong_1      | [INFO] Kong version.......0.3.2
kong_1      |        Proxy HTTP port....8000
kong_1      |        Proxy HTTPS port...8443
kong_1      |        Admin API port.....8001
kong_1      |        dnsmasq port.......8053
kong_1      |        Database...........cassandra keepalive=60000 hosts=cassandra port=9042 timeout=1000 keyspace=kong
kong_1      | [INFO] Connecting to the database...
kong_1      | [ERR] Cassandra error: connection refused
cassandra_1 | INFO  06:33:10 Hostname: ff52d03a1055
cassandra_1 | INFO  06:33:10 Loading settings from file:/etc/cassandra/default.conf/cassandra.yaml
cassandra_1 | INFO  06:33:10 Node configuration:[authenticator=AllowAllAuthenticator; authorizer=AllowAllAuthorizer; auto_snapshot=true; batch_size_warn_threshold_in_kb=5; batchlog_replay_throttle_in_kb=1024; broadcast_rpc_address=127.0.0.1; cas_contention_timeout_in_ms=1000; client_encryption_options=<REDACTED>; cluster_name=Test Cluster; column_index_size_in_kb=64; commit_failure_policy=stop; commitlog_directory=/var/lib/cassandra/commitlog; commitlog_segment_size_in_mb=32; commitlog_sync=periodic; commitlog_sync_period_in_ms=10000; compaction_throughput_mb_per_sec=16; concurrent_counter_writes=32; concurrent_reads=32; concurrent_writes=32; counter_cache_save_period=7200; counter_cache_size_in_mb=null; counter_write_request_timeout_in_ms=5000; cross_node_timeout=false; data_file_directories=[/var/lib/cassandra/data]; disk_failure_policy=stop; dynamic_snitch_badness_threshold=0.1; dynamic_snitch_reset_interval_in_ms=600000; dynamic_snitch_update_interval_in_ms=100; endpoint_snitch=SimpleSnitch; hinted_handoff_enabled=true; hinted_handoff_throttle_in_kb=1024; incremental_backups=false; index_summary_capacity_in_mb=null; index_summary_resize_interval_in_minutes=60; inter_dc_tcp_nodelay=false; internode_compression=all; key_cache_save_period=14400; key_cache_size_in_mb=null; listen_address=localhost; max_hint_window_in_ms=10800000; max_hints_delivery_threads=2; memtable_allocation_type=heap_buffers; native_transport_port=9042; num_tokens=256; partitioner=org.apache.cassandra.dht.Murmur3Partitioner; permissions_validity_in_ms=2000; range_request_timeout_in_ms=10000; read_request_timeout_in_ms=5000; request_scheduler=org.apache.cassandra.scheduler.NoScheduler; request_timeout_in_ms=10000; row_cache_save_period=0; row_cache_size_in_mb=0; rpc_address=0.0.0.0; rpc_keepalive=true; rpc_port=9160; rpc_server_type=sync; saved_caches_directory=/var/lib/cassandra/saved_caches; seed_provider=[{class_name=org.apache.cassandra.locator.SimpleSeedProvider, parameters=[{seeds=127.0.0.1}]}]; server_encryption_options=<REDACTED>; snapshot_before_compaction=false; ssl_storage_port=7001; sstable_preemptive_open_interval_in_mb=50; start_native_transport=true; start_rpc=true; storage_port=7000; thrift_framed_transport_size_in_mb=15; tombstone_failure_threshold=100000; tombstone_warn_threshold=1000; trickle_fsync=false; trickle_fsync_interval_in_kb=10240; truncate_request_timeout_in_ms=60000; write_request_timeout_in_ms=2000]
dockerkong_kong_1 exited with code 1
Gracefully stopping... (press Ctrl+C again to force)
Stopping dockerkong_cassandra_1...

What am I doing wrong here?

P.s. I have a feeling that Cassandra is still booting (not yet listening) and Kong immediately shuts down if it can not connect to Cassandra right away...?

When overriding configuration, do I override the entire kong.yml file with a complete config, or can I just include the lines that I wish to override?

With the current Docker file ( 0.6.0 ), you can't create a container in the background -d, stop it and restart it. Kong always complains about dnsmasq is still running.

I'm pretty certain it is the way the default CMD is set

CMD kong start && tail -f /usr/local/kong/logs/error.log

This makes tail the command that docker is interacting with when you run docker stop so kong never shuts down properly.

I think the default kong.yml should set nginx daemon on and kong start should be set as the command.

Trying to set Kong up in a docker-compose environment.

Here is my docker-compose.yml:

version: '2'
services:
  kong-cassandra:
    image: cassandra:2.2.4
    volumes:
      - ./cassandra:/var/lib/cassadra
    ports:
      - 9042

  kong:
    image: mashape/kong:0.7.0
    restart: always
    ports:
      - 8000:8000
      - 8001:8001
      - 8443:8443
      - 7946:7946
      - 7946:7946/udp
    links:
      - kong-cassandra:cassandra
      - lb

  lb:
    image: dockercloud/haproxy
    links:
      - serviceA
    environment:
      - DOCKER_TLS_VERIFY
      - DOCKER_HOST
      - DOCKER_CERT_PATH
    volumes:
      - $DOCKER_CERT_PATH:$DOCKER_CERT_PATH
    ports:
      - 80

  serviceA:
    extends:
      file: common.yml
      service: microservice
    build:
      context: services/serviceA/
    environment:
      - 'VIRTUAL_HOST=*/serviceA, */serviceA/*'
    command: "python main.py"

After starting up, the Kong container eats up more and more memory up to about 1.2 GB, and keeps both cores in the docker-machine at high usage.

At first, I am not consistently able to GET on ports 8000 and 8001. Sometimes they are available after a fresh startup, but after a while those endpoints start timing out. I have not yet been able to set it up so that requests are passed on to the load balancer behind it.

CONTAINER           CPU %               MEM USAGE / LIMIT   MEM %               NET I/O               BLOCK I/O
3ccd4130717e        0.02%               19.01 MB / 2.1 GB   0.91%               21.91 kB / 8.606 kB   2.822 MB / 0 B
687d13e230bf        178.61%             1.18 GB / 2.1 GB    56.19%              15.4 kB / 17.62 kB    16.38 kB / 106.5 kB
9b4b4fce624a        0.43%               37.04 MB / 2.1 GB   1.76%               22.05 kB / 8.596 kB   4.19 MB / 0 B
bc4f15fcbd70        0.02%               24.94 MB / 2.1 GB   1.19%               103.5 kB / 60.59 kB   7.692 MB / 8.192 kB
c3774ed1f6f7        0.80%               455.5 MB / 2.1 GB   21.69%              40.34 kB / 30.6 kB    51.06 MB / 1.004 MB
e6b19df2c5cb        0.02%               23.33 MB / 2.1 GB   1.11%               22.18 kB / 8.638 kB   7.442 MB / 0 B```

Will edit as I hack away...

Hello!

This is kind of a documentation problem/FYI, so just if someone else bumps into this, don't get too surprised.

Using docker-engine 1.10.1-0~wily (so the ubuntu docker package) results in a strange error:

$ docker run mashape/kong /usr/local/bin/serf agent 
==> Starting Serf agent...
==> Failed to start the Serf agent: Error creating Serf: Failed to create memberlist: Failed to start TCP listener. Err: listen tcp 0.0.0.0:7946: socket: operation not permitted

So one has to start it with a relaxed --security-opt setting.

$ docker run -it --security-opt seccomp:unconfined  mashape/kong /usr/local/bin/serf agent 
==> Starting Serf agent...
==> Starting Serf agent RPC...
==> Serf agent running!
         Node name: 'f8e0492ca4ff'
         Bind addr: '0.0.0.0:7946'
          RPC addr: '127.0.0.1:7373'
         Encrypted: false
          Snapshot: false
           Profile: lan

==> Log data will now stream in as it occurs:

    2016/02/15 10:18:46 [INFO] agent: Serf agent starting
    2016/02/15 10:18:46 [INFO] serf: EventMemberJoin: f8e0492ca4ff 172.17.0.2
    2016/02/15 10:18:47 [INFO] agent: Received event: member-join

^C

==> Caught signal: interrupt
==> Gracefully shutting down agent...
    2016/02/15 10:18:47 [INFO] agent: requesting graceful leave from Serf
    2016/02/15 10:18:47 [INFO] serf: EventMemberLeave: f8e0492ca4ff 172.17.0.2
    2016/02/15 10:18:47 [INFO] agent: requesting serf shutdown
    2016/02/15 10:18:47 [INFO] agent: shutdown complete

https://docs.docker.com/docker-hub/official_repos/

• Git repository
  • Hosted on GitHub
  • Public
• Dockerfile
  • Written FROM scratch or another Official Image.
  • Follow Dockerfile best practises (1, 2)
• Short description
  • named README-short.txt
  • in the git repo for the latest tag
  • no more than 100 characters
• Logo
  • named logo.png
  • in the git repo for the latest tag
  • 200px square, or max-width (preferably square or landscape)
• Long description
  • named README.md
  • in the git repo for the latest tag
  • Content
    • Overview & links (What is ...?)
      • Overview of the software
      • Selection of relevant links
      • mandatory link to the Dockerfile
    • How-to/Usage (How to use...)
      • Describe how to run the image, common use cases and examples. Step-by-step instructions.
        • Link with Cassandra container
        • Custom configuration
        • Example command: kong reload
    • Issues & contributions
      • Links to contribute to the project
      • Contribution guidelines
      • A link to Docker’s resources for contributors
      • Where they can get help
        • Gitter
        • Google groups
        • GitHub Issues
• License
  • License file in the git repo, latest tag

Hey,

I sometimes have issues with kong and I get the msg:

Run with --v (verbose) or --vv (debug)

The Dockerfile has the command

CMD ["kong", "start"]

I am creating my own image because i need to put plugins code into kong, so I have something like this:

FROM kong:0.9.5

COPY ...
COPY ...
COPY ...
    
ENV DATABASE postgres 

Now.. is there a way for me to start kong in debug mode or do I have clone this repo and change the cmd to the debug mode?

If there is no way could some dev put the kong start in the docker-entrypoint.sh? I can override that in my image without issues.

Thanks

ideally host/port/user/database/password (or at bare minimum password) should be supplied via environment variables (or whatever other way), but it should not be left out to defaults without any option to change it