From the docker linking
I can have A container links to B container.
Then I can see the B's ip address and exposed port in A's ENV variables.
However, how can I figure out A's ip address wihtin B container?
To find one container from another, you can use a 'service discovery' mechanism such as SkyDock.
Skydock - Automagic Service Discovery for Docker
Skydock monitors docker events when containers start, stop, die, kill, etc and inserts records into a dynamic DNS server skydns. This allows standard DNS queries for services running inside docker containers.
For the more complex case where your containers are on multiple hosts and you need a way to network them together, see weave-dns (Please note I work on weave and weave-dns).
Related
I have a set of docker containers created with docker compose, which creates a "user-defined" bridge network.
One of these docker containers is running DNSmasq so we can define custom (internal) domain names to point to local IPs.
Trouble is, none of the other docker containers can resolve these domain names. I think the issue is that I can't get the docker DNS to forward its requests to the docker container running DNSmasq (i.e. it doesn't even know it exists).
As a test, I did a docker network inspect <network created by docker compose> and noted the IP address of the DNSmasq container. Then, in one of the other containers' /etc/resolv.conf, I set nameserver to that IP address. Then I can resolve all these internal domains names.
Sadly, putting dnsmasq in there doesn't work, despite the fact that the user-defined network has automatic service discovery enabled.
It seems that one way to make this work then is to force my DNSmasq container to always have the same IP, and then make sure the other docker containers point to that as their nameserver, e.g. by defining the network explicitly in the compose file.
Is there no other way? I'd rather not have to define the entire network to replicate this automatically created one when all I want is to know the IP address of one container.
I am building a microservices based application and would like to use Consul as service registry. All in all I have three scenarios:
All the services run on the host.
All the services run on the host, but Consul runs in Docker.
All the services and Consul run in Docker.
Now I have the problem of how to register the services with their IP address, because I need to figure out their IP address so that it is reachable by Consul (e.g., for the health checks):
If everything runs on the same host, it's pretty easy: Simply use 127.0.0.1, and you're done.
If everything (including Consul) runs in Docker, I could use hostname -i from within the Docker containers to figure out their external IP and hand it over to Consul. This works, but I wonder if there is a better way to solve this? (Ideally, the solution should also work in the same way on Kubernetes.)
If the services run on the host, but Consul runs in Docker, right now I am missing any idea at all. Basically, Consul requires the host's IP address to be able to talk to the services, but I can only detect this from within the Consul container (by resolving host.docker.internal). But first, this does not work from externally, and second it only works for Docker for Mac / Windows, not e.g. with Kubernetes.
How could I solve these issues?
PS: I would like to avoid using a container such as registrator by Gliderlabs, since I have doubts how well this works on Kubernetes, and also it won't help with the mixed Docker / host scenario.
If you're using Kubernetes, you might start by checking whether its built-in service registry meets your needs. There's generally not a direct path to reach a pod via its node's host's IP address, so the setup you describe won't really work well. (I might consider Consul for a key/value store but I wouldn't reach for it as a service registry in Kubernetes land.)
In plain multi-host Docker land, this is one of the few situations I've found where host networking is appropriate. Start Consul with --net host or an equivalent option in Docker Compose or another orchestration tool. Then Consul will believe "its" IP address is the host's, and if you have automated TCP probes of well-known ports, you can search every service that's running on the host and discover e.g. a MySQL service on port 3306, whether running in a container or natively on the host.
With this setup, servicename.service.consul will resolve to some physical-host IP address. If you have a Docker container pointing at its current host for DNS service, then that will route a service to some host, maybe the same one, but this has worked reliably for me in the past.
Note that the relevant hostnames will be different in different environments: servicename.service.consul for a Consul-based setup, servicename.namespacename.svc.cluster.local in Kubernetes, maybe localhost in a developer-desktop environment. You need to make sure this is configurable, most straightforwardly via an environment variable.
I have two instances of keycloak running on container each on is running on a single node.
The nodes are bare-metal nodes inside my company network.
keycloak uses TCPPING as discovery protocol.
Since the two containers are running on different nodes, and each instance is pining inside docker default network they are not able to find each other.
I said docker default network because I didn’t specify special network for the two containers.
Any idea how can I make the two instances in this architectural design discover each others!
and I was thinking about docker swarm as a solution.
Assuming the two nodes are on the same network and are able to connect to each other, you can get the two container to discover each other using docker host networking
It would be as easy as docker run --net=host
Docker host networking makes the container to use the networking of the host node and thus will be allocated an IP address by the DHCP server used by the host node and for all practical purposes , would look like another host in that network.
This allows the two containers to discover each other using TCPPING
Docker swarm would also enable this .Docker swarm basically abstracts multiple host nodes such that you can containers on them as if you are running docker on single host. But that will require docker-machine and whole new setup.
How do you access remote Docker container by its hostname?
I need to access remote Docker containers by its hostnames (or some constant IP's) for development and testing purposes. I have tried:
looking for any DNS approach (have not found any clues),
importing /ets/hosts (probably impossible),
creating tunnes (only this works but it is very time consuming).
It's the same as running any other process on a host, Docker or not Docker: you access it via the host name or IP address of the host and the port the service is listening on (the first port of the docker run -p argument). Docker containers don't have externally visible individual IP addresses any more than non-Docker HTTP or ssh daemons do.
If you do have DNS infrastructure available to you, you could set up CNAME records to resolve particular service names to the specific hosts that are running them.
One solution that may help you is some sort of service registry; in the past I've used Consul with some success. You can configure Consul with some health checks or other probes ("look for an HTTP service on port 12345 that answers GET / calls"), and it will provide its own DNS service ("okay, http://whatevername.service.consul:12345/ will reach your service on whichever hosts it happens to be running on").
Nothing in the Docker infrastructure specifically helps this. Using /etc/hosts is distinctly not a best practice: the name-to-IP mapping needs to be kept in sync across all machines and you'll start wishing you had a network service to publish it for you, which is exactly what DNS is for.
Elasticsearch is designed to run in cluster mode, all I have to do is to define the relevant node IPs in the cluster via environment variable and as long as network connectivity is available it will connect and join the other nodes to the cluster.
I have 3 nodes, 1 is acting as the docker swarm manager and the other two are workers. I have initialized the manager and joined the worker nodes and everything looks ok from that standpoint.
Now I'm trying to run the elasticsearch container in a way that will allow me to join all nodes to the same elasticsearch cluster, however, I want the nodes to join using their overlay network interface and that means that I need to know the container internal IP addresses at the time of running the docker service create command, how can I do this? Do I have to use something like consul to achieve this?
Some clarifications:
I need to know, at the time of service creation the IP addresses (or DNS names) for all Elasticsearch participants so I could start the cluster correctly. This has to be at the time of creation and not afterwards. Also, as I understand, I can expose ports 9200/9300 for all services and work with the external machine IPs and get it to work, but I would like to use the overlay network to do all these communications (I thought this is what swarm mode is for).
Only a partial solution here.
So, when attaching your services to a custom overlay network, you indeed have access to Docker's custom Service discovery feature. I'll detail the networking feature of Docker Swarm mode, before trying to tie it to your problem.
I'll be using the different term of services and tasks, in which a service could be elasticsearch, whereas a task is a single instance of that elasticsearch service.
Docker networking
The idea is that for each services you create, docker assigns a Virtual IP (VIP), and a custom dns alias. You can retrieve this VIP using the docker service inspect myservice command.
But, there is two modes to attach a service to an overlay network dnsrr and VIP. You can select these options using the --endpoint-mode options of docker service create.
The VIP mode (I believe it is the default one, or at least the most used), affects the virtual ip to the service's dns alias. This means that doing an nslookup servicename would return to you a single vip, that behind the scenes, would be linked to one of your container in a round robin fashion. But, there is also a special dns alias that lets you access all of your instances ips (all of your tasks ips) : tasks.myservice.
So in VIP mode you can retrieve all of your tasks ips using a simple nslookup tasks.myservice, where myservice is a service name.
The other mode is dnsrr. This mode simply gets rid of the VIP, and connects the dns alias to the different tasks (=service instances), in a round robin way. This way, you simply have to do a nslookup myservice to retrieve the different service instances ip.
Elasticsearch clustering
Ok so first of all I'm not really familiar with the way elasticsearch lets you cluster. From what I understood from your question, you need when running the elasticsearch binary, give it as a parameter, the adress of all of the other nodes it needs to cluster with.
So what I would do, is to create a custom Elasticsearch image, probably based on the one from the default library, to which I would add a custom Entrypoint that would firstly run a script to retrieve the other tasks ip.
I'd believe that staying in VIP mode is suitable for you, since there is the tasks.myservice dns alias. You'll then need to parse the output to retrieve the tasks ip (and probably remove yours). Then you'll be able to save them in a config file environment variable, or use them as a runtime option for your elasticsearch binary.
Edit: To create a custom overlay network, you will need to use the docker network create command, and use the --network option of docker service create
This is answer is mainly based on the Swarm mode networking documentation