Get Certified: Proven Methods to Pass the Docker DCA Exam

= The command docker service create --network --secure will not ensure that overlay traffic between service tasks is encrypted. This is because the --secure option is not a valid option for the docker service create command1. To ensure that overlay traffic between service tasks is encrypted, you need to use the --opt encrypted option when creating the overlay network with the docker network create command2. For example, to create an encrypted overlay network named my-net, you can use the following command:

docker network create --driver overlay --opt encrypted my-net

Then, you can use the --network my-net option when creating the service with the docker service create command3. For example, to create a service named my-service using the nginx image and the my-net network, you can use the following command:

docker service create --name my-service --network my-net nginx

References:

• docker service create | Docker Docs
• Use overlay networks | Docker Docs
• Create a service | Docker Docs

The networking drivers that allow you to enable multi-host network connectivity between containers are bridge, macvlan, ipvlan, and overlay. These drivers create networks that can span multiple Docker hosts, and therefore enable containers on different hosts to communicate with each other. The other drivers, such as host, user-defined, and none, create networks that are either isolated or limited to a single host. Here is a brief overview of each driver and how it supports multi-host networking:

•bridge: The bridge driver creates a network that connects containers on the same host using a Linux bridge. However, it can also be used to create a network that connects containers across multiple hosts using an external key-value store, such as Consul, Etcd, or ZooKeeper. This feature is deprecated and not recommended, as it requires manual configuration and has some limitations. The preferred driver for multi-host networking is overlay1.

•macvlan: The macvlan driver creates a network that assigns a MAC address to each container, making it appear as a physical device on the network. This allows the containers to communicate with other devices on the same network, regardless of the host they are running on. The macvlan driver can also use 802.1q trunking to create sub-interfaces and isolate traffic between different networks2.

•ipvlan: The ipvlan driver creates a network that assigns an IP address to each container, making it appear as a logical device on the network. This allows the containers to communicate with other devices on the same network, regardless of the host they are running on. The ipvlan driver can also use different modes, such as l2, l3, or l3s, to control the routing and isolation of traffic between different networks3.

•overlay: The overlay driver creates a network that connects multiple Docker daemons together using VXLAN tunnels. This allows the containers to communicate across different hosts, even if they are on different networks. The overlay driver also supports encryption, load balancing, and service discovery. The overlay driver is the default and recommended driver for multi-host networking, especially for Swarm services4.

References:

•Use bridge networks

•Use macvlan networks

•Use ipvlan networks

•Use overlay networks

The answer depends on whether you want to access the container from the host’s network or from other containers on the same network. EXPOSE and --publish have different effects on the container’s port visibility. References: Docker run reference, Dockerfile reference, Docker networking overview

= This is not how the seven managers should be distributed across three datacenters or availability zones. A swarm cluster is a group of Docker hosts that are running in swarm mode and act as managers or workers1. A manager node is responsible for maintaining the swarm state and orchestrating the services2. A swarm cluster needs a quorum of managers to operate, which means a majority of managers must be available and able to communicate with each other3.

The problem with distributing the seven managers as 4-2-1 is that it creates a split-brain scenario, where the cluster can lose the quorum if one datacenter or availability zone fails. For example, if the datacenter with four managers goes down, the remaining three managers will not have enough votes to form a quorum, and the cluster will stop functioning. Similarly, if the datacenter with one manager goes down, the cluster will lose the tie-breaking vote and will not be able to elect a leader4.

A better way to distribute the seven managers across three datacenters or availability zones is to use 3-2-2, which ensures that the cluster can tolerate the failure of any one datacenter or availability zone and still maintain the quorum. For example, if the datacenter with three managers goes down, the remaining four managers will have enough votes to form a quorum and elect a leader. Similarly, if the datacenter with two managers goes down, the remaining five managers will have enough votes to form a quorum and elect a leader4. References:

• Swarm mode overview | Docker Docs
• Administer and maintain a swarm of Docker Engines | Docker Docs
• Raft consensus in swarm mode | Docker Docs
• Docker Swarm: How to distribute managers across availability zones? - Stack Overflow

The command docker run -v /data:/mydata -mode readonly ubuntu will not mount the host’s /data1 directory to the ubuntu container in read-only mode. The command has several errors that prevent it from working correctly. First, the host directory should be /data1 instead of /data, as specified in the question. Second, the option flag should be --mode instead of -mode, and it should be placed before the image name. Third, the mode value should be ro instead of readonly, as per the Docker documentation1. The correct command should be docker run -v /data1:/mydata --mode ro ubuntu, which will mount the host’s /data1 directory as a read-only volume at /mydata inside the container1. References:

• docker run | Docker Docs

Yes, the docker port command can be used to identify the published ports for a running container. It shows the mapping between the host ports and the container’s exposed ports. The docker network inspect command can also provide information about the network settings of the container, including port mappings. However, it’s important to note that docker network inspect requires the network’s name or ID as an argument, not the container’s. Therefore, to get the network details of a specific container, you would first need to identify the network the container is connected to. These commands, when used appropriately, can help you identify the published ports for a container.

= A DTR security scan will detect private keys copied to the image. DTR security scan is a feature of Docker Trusted Registry (DTR) that scans images to detect any security vulnerability1. DTR security scan uses the open source tool SecretScanner2 to find unprotected secrets in container images or file systems. SecretScanner can match the contents of images against a database of approximately 140 secret types, including private keys3. Therefore, if an image contains private keys, DTR security scan will report them as potential secrets and alert the user to remove them from the image. References:

• Scan images for vulnerabilities | Docker Docs
• GitHub - deepfence/SecretScanner: :unlock: Find secrets and passwords …
• SecretScanner/deepfence_secret_scanner.py at main · deepfence/SecretScanner

Using either EXPOSE or -publish to access the container on the bridge network will not accomplish the goal of creating a container that is reachable from its host’s network. EXPOSE is a way of documenting which ports a container listens on, but it does not open any ports to the host1. -publish (or -p) is a way of mapping a host port to a container port, but it does not change the network mode of the container2. By default, Docker containers use the bridge network, which isolates them from the host network3. To create a container that is reachable from its host’s network, you need to use the --network host option when running the container4. This will make the container share the host’s network stack and have the same IP address as the host4. References:

• 1: Difference Between “expose” and “publish” in Docker | Baeldung on Ops
• 2: Deploy services to a swarm | Docker Docs
• 3: Bridge network | Docker Docs
• 4: Host network | Docker Docs

 = Host is not a type of Linux kernel namespace that provides container isolation. Linux namespaces are a feature of the Linux kernel that partitions kernel resources such that one set of processes sees one set of resources while another set of processes sees a different set of resources1. There are eight kinds of namespaces available: Mount, Process, User, Network, UTS, IPC, Cgroup, and Time1. Host is a parameter that can be used to run a container in the host’s network namespace, which means the container shares the same network interfaces and configuration as the host2. References:

• Linux namespaces - Wikipedia
• Network settings | Docker Documentation

= The pid namespace is not a Linux kernel namespace that is disabled by default and must be enabled at Docker engine runtime to be used. The pid namespace is one of the six namespaces that are enabled by default when you run a container with Docker1. The pid namespace isolates the process ID number space, meaning that processes in different pid namespaces can have the same PID2. This allows containers to have their own init process with PID 1 and to limit the visibility and interaction of processes between containers and the host3. To disable the pid namespace, you need to use the --pid option with the docker run command and specify the host or another container as the pid mode. References:

• Docker run reference | Docker Docs
• pid_namespaces(7) - Linux manual page - man7.org
• Building containers by hand: The PID namespace - Enable Sysadmin
• [Share host and container processes with --pid | Docker Docs]