Initialize swarm mode:
root#ip-172-31-44-207:/home/ubuntu# docker swarm init --advertise-addr 172.31.44.207
Swarm initialized: current node (4mj61oxcc8ulbwd7zedxnz6ce) is now a manager.
To add a worker to this swarm, run the following command:
Join the second node:
docker swarm join \
--token SWMTKN-1-4xvddif3wf8tpzcg23tem3zlncth8460srbm7qtyx5qk3ton55-6g05kuek1jhs170d8fub83vs5 \
172.31.44.207:2377
To add a manager to this swarm, run 'docker swarm join-token manager' and follow the instructions.
# start 2 services
docker service create continuumio/miniconda3
docker service create --name redis redis:3.0.6
root#ip-172-31-44-207:/home/ubuntu# docker service ls
ID NAME REPLICAS IMAGE COMMAND
2yc1xjmita67 miniconda3 0/1 continuumio/miniconda3
c3ptcf2q9zv2 redis 1/1 redis:3.0.6
As shown above, redis has it's replica while miniconda does not seem to be replicated.
I do usually log-in to miniconda container to type these commands:
/opt/conda/bin/conda install jupyter -y --quiet && mkdir /opt/notebooks && /opt/conda/bin/jupyter notebook --notebook-dir=/opt/notebooks --ip='*' --port=8888 --no-browser
The problem is that docker exec -it XXX bash command does not work with swarm mode.
You can execute commands by filtering container name without needing to pass the entire swarm container hash, just by the service name. Like this:
docker exec $(docker ps -q -f name=servicename) ls
There is one liner for accessing corresponding instance of the service for localhost:
docker exec -ti stack_myservice.1.$(docker service ps -f 'name=stack_myservice.1' stack_myservice -q --no-trunc | head -n1) /bin/bash
It is tested on PowerShell, but bash should be the same. The oneliner accesses the first instance, but replace '1' with the number of the instance you want to access in two places to get other one.
More complex example is for distributed case:
#! /bin/bash
set -e
exec_task=$1
exec_instance=$2
strindex() {
x="${1%%$2*}"
[[ "$x" = "$1" ]] && echo -1 || echo "${#x}"
}
parse_node() {
read title
id_start=0
name_start=`strindex "$title" NAME`
image_start=`strindex "$title" IMAGE`
node_start=`strindex "$title" NODE`
dstate_start=`strindex "$title" DESIRED`
id_length=name_start
name_length=`expr $image_start - $name_start`
node_length=`expr $dstate_start - $node_start`
read line
id=${line:$id_start:$id_length}
name=${line:$name_start:$name_length}
name=$(echo $name)
node=${line:$node_start:$node_length}
echo $name.$id
echo $node
}
if true; then
read fn
docker_fullname=$fn
read nn
docker_node=$nn
fi < <( docker service ps -f name=$exec_task.$exec_instance --no-trunc -f desired-state=running $exec_task | parse_node )
echo "Executing in $docker_node $docker_fullname"
eval `docker-machine env $docker_node`
docker exec -ti $docker_fullname /bin/bash
This script could be used later as:
swarm_bash stack_task 1
It just execute bash on required node.
EDIT 2017-10-06:
Nowadays you can create the overlay network with --attachable flag to enable any container to join the network. This is great feature as it allows a lot of flexibility.
E.g.
$ docker network create --attachable --driver overlay my-network
$ docker service create --network my-network --name web --publish 80:80 nginx
$ docker run --network=my-network -ti alpine sh
(in alpine container) $ wget -qO- web
<!DOCTYPE html>
<html>
<head>
....
You are right, you cannot run docker exec on docker swarm mode service. But you can still find out, which node is running the container and then run exec directly on the container. E.g.
docker service ps miniconda3 # find out, which node is running the container
eval `docker-machine env <node name here>`
docker ps # find out the container id of miniconda
docker exec -it <container id here> sh
In your case you first have to find out, why service cannot get the miniconda container up. Maybe running docker service ps miniconda3 shows some helpful error messages..?
Using the Docker API
Right now, Docker does not provide an API like docker service exec or docker stack exec for this. But regarding this, there already exists two issues dealing with this functionality:
github.com - moby/moby - Docker service exec
github.com - docker/swarmkit - Support for executing into a task
(Regarding the first issue, for me, it was not directly clear that this issue deals with exactly this kind of functionality. But Exec for Swarm was closed and marked as duplicate of the Docker service exec issue.)
Using Docker daemon over HTTP
As mentioned by BMitch on run docker exec from swarm manager, you could also configure the Docker daemon to use HTTP and than connect to every node without the need of ssh. But you should protect this using TLS authentication which is already integrated into Docker. Afterwards you would be able to execute the docker exec like this:
docker --tlsverify --tlscacert=ca.pem --tlscert=cert.pem --tlskey=key.pem \
-H=$HOST:2376 exec $containerId $cmd
Using skopos-plugin-swarm-exec
There exists a github project which claims to solve the problem and provide the desired functionality binding the docker daemon:
docker run -v /var/run/docker.sock:/var/run/docker.sock \
datagridsys/skopos-plugin-swarm-exec \
task-exec <taskID> <command> [<arguments>...]
As far as I can see, this works by creating another container at same node where the container reside where the docker exec should by executed on. On this node this container mounts the docker daemon socket to be able to execute docker exec there locally.
For more information have a look at: skopos-plugin-swarm-exec
Using docker swarm helpers
There is also another project called docker swarm helpers which seems to be more or less a wrapper around ssh and docker exec.
Reference:
https://github.com/docker/swarmkit/issues/1895#issuecomment-302147604
https://github.com/docker/swarmkit/issues/1895#issuecomment-358925313
You can jump in a Swarm node and list the docker containers running using:
docker container ls
That will give you the container name in a format similar to: containername.1.q5k89uctyx27zmntkcfooh68f
You can then use the regular exec option to run commands on it:
docker container exec -it containername.1.q5k89uctyx27zmntkcfooh68f bash
created a small script for our docker swarm cluster.
this script takes 3 params. first is the service you want to connect to second the task you want to run this can be /bin/bash or any other process you want to run. Third is optional and will fill -c option for bash or sh
-n is optional to force it to connect to a node
it retrieves the node that runs the service and runs the command.
#! /bin/bash
set -e
task=${1}
service=$2
bash=$3
serviceID=$(sudo docker service ps -f name=$service -f desired-state=running $service -q --no-trunc |head -n1)
node=$(sudo docker service ps -f name=$service -f desired-state=running $service --format="{{.Node}}"| head -n1 )
sudo docker -H $node exec -it $service".1."$serviceID $bash -c "$task"
note: this requires the docker nodes to accept tcp connections by exposing docker on port 2375 on the worker nodes
For those who have multiple replicas and just want to run a command within any of them, here is another shortcut:
docker exec -it $(docker ps -q -f name=SERVICE_NAME | head -1) bash
I wrote script to exec command in docker swarm by service name. For example it can be used in cron. Also you can use bash pipelines and passes all params to docker exec command. But works only on same node where service started. I wish it could help someone
#!/bin/bash
# swarm-exec.sh
set -e
for ((i=1;i<=$#;i++)); do
val=${!i}
if [ ${val:0:1} != "-" ]; then
service_id=$(docker ps -q -f "name=$val");
if [[ $service_id == "" ]]; then
echo "Container $val not found!";
exit 1;
fi
docker exec ${#:1:$i-1} $service_id ${#:$i+1:$#};
exit 0;
fi
done
echo "Usage: $0 [OPTIONS] SERVICE_NAME COMMAND [ARG...]";
exit 1;
Example of using:
./swarm-exec.sh app_postgres pg_dump -Z 9 -F p -U postgres app > /backups/app.sql.gz
echo ls | ./swarm-exec.sh -i app /bin/bash
./swarm-exec.sh -it some_app /bin/bash
The simpliest command I found to docker exec into a swarm node (with a swarm manager at $SWARM_MANAGER_HOST) running the service $SERVICE_NAME (for example mystack_myservice) is the following:
SERVICE_JSON=$(ssh $SWARM_MANAGER_HOST "docker service ps $SERVICE_NAME --no-trunc --format '{{ json . }}' -f desired-state=running")
ssh -t $(echo $SERVICE_JSON | jq -r '.Node') "docker exec -it $(echo $SERVICE_JSON | jq -r '.Name').$(echo $SERVICE_JSON | jq -r '.ID') bash"
This asserts that you have ssh access to $SWARM_MANAGER_HOST as well as the swarm node currently running the service task.
This also asserts that you have jq installed (apt install jq), but if you can't or don't want to install it and you have python installed you can create the following alias (based on this answer):
alias jq="python3 -c 'import sys, json; print(json.load(sys.stdin)[sys.argv[2].partition(\".\")[-1]])'"
See addendum 2...
Example of a oneliner for entering the database my_db on node master:
DB_NODE_ID=master && docker exec -it $(docker ps -q -f name=$DB_NODE_ID) mysql my_db
In case you want to configure, say max_connections:
DB_NODE_ID=master && $(docker exec -it $(docker ps -q -f name=$DB_NODE_ID) mysql -e "SET GLOBAL max_connections = 1000") && docker exec -it $(docker ps -q -f name=$DB_NODE_ID) mysql my_db
This approach allows to enter all database nodes (e.g. slaves) just by setting the DB_NODE_ID variable accordingly.
Example for slave s2:
DB_NODE_ID=s2 && docker exec -it $(docker ps -q -f name=$DB_NODE_ID) mysql my_db
or
DB_NODE_ID=s2 && $(docker exec -it $(docker ps -q -f name=$DB_NODE_ID) mysql -e "SET GLOBAL max_connections = 1000") && docker exec -it $(docker ps -q -f name=$DB_NODE_ID) mysql my_db
Put this into your KiTTY or PuTTY configuration for master / s2 under Data/Command and you are set.
As an addendum:
The old, non swarm mode version reads simply
docker exec -it master mysql my_db
resp.
DB_ID=master && $(docker exec -it $DB_ID mysql -e "SET GLOBAL max_connections = 1000") && docker exec -it $DB_ID mysql tmp
Addendum 2:
As it turned out by example, the term docker ps -q -f name=$DB_NODE_ID may return wrong values under certain conditions.
The following approach works correctily:
docker ps -a | grep "_$DB_NODE_ID." | awk '{print $1}'
You may substitute the examples above accordingly.
Addendum 3:
Well, these terms look awful and they certainly are painful to type, so you may want to ease your work. On Linux, everybody knows how to do this. On Windws, you may want to use AHK.
This is the AHK term I use:
:*:ii::DB_NODE_ID=$(docker ps -a | grep "_." | awk '{{}print $1{}}') && docker exec -it $id ash{Left 49}
So when I type ii -- which is as simple as it can get -- I get the desired term with the cursor in place and just have to fill in the container name.
I edited the script Brian van Rooijen added above. Because my reputation is to low, I cannot add it
#! /bin/bash
set -e
service=${1}
shift
task="$*"
echo $task
serviceID=$(docker service ps -f name=$service -f desired-state=running $service -q --no-trunc |head -n1)
node=$(docker service ps -f name=$service -f desired-state=running $service --format="{{.Node}}"| head -n1 )
serviceName=$(docker service ps -f name=$service -f desired-state=running $service --format="{{.Name}}"| head -n1 )
docker -H $node exec -it $serviceName"."$serviceID $task
I had the issue that the container didn't exists with the hard coded .1. in the execution.
Take a look at my solution: https://github.com/binbrayer/swarmServiceExec.
This approach is based on Docker Machines. I also created the prototype of the script to call containers asynchronously and as a result simultaneously.
I'm trying to use this image https://hub.docker.com/_/mariadb/ (any version).
I'm using the following to launch the container:
cd maria
docker build -t maria-image .
docker run --name maria maria-image -d -e MYSQL_ALLOW_EMPTY_PASSWORD=1
cd ..
I'm preparing a custom build in case I need to do any future modifications so that lives in maria/Dockerfile with the following:
FROM mariadb:5.5
MAINTAINER ...
EXPOSE 3306
If I do docker ps -a I get status "Exited (2) 5 seconds ago".
Your args appear to be in the wrong order, maria-image should be after all other docker run args:
docker run --name maria -d -e MYSQL_ALLOW_EMPTY_PASSWORD=1 maria-image
The version you ran passed the -d and -e as the command for docker to run. Note that you'll want to first run docker rm -v maria to free the container name for reuse.
I am following this tutorial for service discovery http://jasonwilder.com/blog/2014/07/15/docker-service-discovery
Briefly:
I created an etcd host running at x.y.z.d:4001
docker run -d --name etcd -p 4001:4001 -p 7001:7001 coreos/etcd
Created a backend server running a container at backend_serverip:8000 and docker-register
$ docker run -d -p 8000:8000 --name whoami -t jwilder/whoami
$ docker run --name docker-register -d -e HOST_IP=$(hostname --all-ip-addresses | awk '{print $1}') -e ETCD_HOST=x.y.z.d:4001 -v /var/run/docker.sock:/var/run/docker.sock -t jwilder/docker-register
Created another backend server running a container at backend2_serverip:8000 and docker-register
$ docker run -d -p 8000:8000 --name whoami -t jwilder/whoami
$ docker run --name docker-register -d -e HOST_IP=$(hostname --all-ip-addresses | awk '{print $1}') -e ETCD_HOST=x.y.z.d:4001 -v /var/run/docker.sock:/var/run/docker.sock -t jwilder/docker-register
Created a client running docker-discover and an ubuntu image
$ docker run -d --net host --name docker-discover -e ETCD_HOST=10.170.71.226:4001 -p 127.0.0.1:1936:1936 -t jwilder/docker-discover
When I look at the logs to see if containers are being registered I see teh folowing error
2015/07/09 19:28:00 error running notify command: python /tmp/register.py, exit status 1
2015/07/09 19:28:00 Traceback (most recent call last):
File "/tmp/register.py", line 22, in <module>
backends = client.read("/backends")
File "/usr/local/lib/python2.7/dist-packages/etcd/client.py", line 347, in read
self.key_endpoint + key, self._MGET, params=params, timeout=timeout)
File "/usr/local/lib/python2.7/dist-packages/etcd/client.py", line 587, in api_execute
return self._handle_server_response(response)
File "/usr/local/lib/python2.7/dist-packages/etcd/client.py", line 603, in _handle_ser
etcd.EtcdError.handle(**r)
File "/usr/local/lib/python2.7/dist-packages/etcd/__init__.py", line 184, in handle
raise exc(msg, payload)
etcd.EtcdKeyNotFound: Key not found : /backends
I tried manually creating this directory , I also tried running the containers with privileged option but no luck
The error you are getting is from a bug in the code. The problem is that /backends does not exist in your etcd directory. You can create it yourself by manually by running this:
curl -L http://127.0.0.1:4001/v2/keys/backends -XPUT -d dir=true
Once the directory exists in etcd, you won't get the error anymore.
I created a pull request that fixes the bug and if you want to use the fixed code, you can build your own image:
git clone git#github.com:rca/docker-register.git
cd docker-register
docker build -t docker-register .
Then your command for docker register would look like:
$ docker run --name docker-register -d -e HOST_IP=$(hostname --all-ip-addresses | awk '{print $1}') -e ETCD_HOST=x.y.z.d:4001 -v /var/run/docker.sock:/var/run/docker.sock -t docker-register
Note I simply removed jwilder/ from the image name in the command so it uses your local version.
I have deis(1.5.2) with 3 host and I want "app" with database. I want to use postgres, so I found this docker image https://registry.hub.docker.com/_/postgres/ . I did deploy without problems, but I don't know how can I connect into this app/container (create some db, users) and link with other app/container. They write commands for it but it's for docker. So how can I run these commands from deis:
docker run --name some-app --link some-postgres:postgres -d application-that-uses-postgres
docker run -it --link some-postgres:postgres --rm postgres sh -c 'exec psql -h "$POSTGRES_PORT_5432_TCP_ADDR" -p "$POSTGRES_PORT_5432_TCP_PORT" -U postgres'
or do you have some other solution for using DB with deis?
I am running the example installation using docker container from the CKAN site. A clean run, that downloads fresh (no local images) gives no errors
docker run -d --name db ckan/postgresql
docker run -d --name solr ckan/solr
docker run -d -p 80:80 --link db:db --link solr:solr ckan/ckan
but then a "docker ps" does not have the ckan image running ...
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
90c6e6a77b0a ckan/solr:latest "java -jar start.jar 15 minutes ago Up 15 minutes 8983/tcp solr
53f9a9f5c145 ckan/postgresql:latest "/usr/local/bin/run" 20 minutes ago Up 20 minutes 5432/tcp db
Where would errors show?
docker ps only lists running containers. You have to pass in the -a option in order to list all containers you have run
docker ps -a
Find the right container and run the following command to see any error messages it may have reported before shutting down
docker logs <container_id>
As at March 6 2015 described at https://github.com/ckan/ckan/issues/2255
ensure that you have killed off any solr containers still running using
docker rm <idnumber>
then use the more recent schema config
wget https://github.com/ckan/ckan/blob/master/ckan/config/solr/schema.xml
docker run -d --name solr -v `pwd`/schema.xml:/opt/solr/example/solr/ckan/conf/schema.xml ckan/solr
then repeat the original failing step:
docker run -d -p 80:80 --link db:db --link solr:solr ckan/ckan
-- EDIT -- This looks like it works but it introduces other problems.
Am going to experiment with suggestion to use https://github.com/datacats/datacats