Q: A security audit highlights that Docker's default seccomp profile allows over 300 system calls, including dangerous syscalls like `ptrace`, `bpf`, `keyctl`, and `perf_event_open` that have historically been exploited for kernel privilege escalation. For your public-facing Nginx reverse proxy and payment processing microservices, you must craft strict, least-privilege seccomp profiles that block unnecessary kernel interfaces, and deploy an AppArmor profile preventing unauthorized file access and raw socket creation.
Protect host kernels against zero-day container breakouts by implementing custom seccomp BPF syscall whitelisting filters and granular AppArmor mandatory access control profiles.
Want to master this scenario in a live sandbox? KodeKloud's Docker Certified Associate (DCA) Hands-On Lab Course covers this exact problem with hands-on terminal drills.
🛠️ Production Runbook & Step-by-Step Resolution
Direct Answer: Protect host kernels against zero-day container breakouts by implementing custom seccomp BPF syscall whitelisting filters and granular AppArmor mandatory access control profiles.
Audit Application System Calls via strace and eBPF
Trace the actual system calls executed by your microservice during startup and peak simulated workload using strace or BCC/bpftrace syscount to establish an accurate whitelist baseline.
# Trace all unique system calls executed by the application
strace -c -f -p $(docker inspect --format '{{.State.Pid}}' web-app) -- sleep 30
# Or run application under strace to capture bootstrap syscalls
docker run --rm web-app:latest strace -ff -e trace=all /app/binary 2>&1 | awk -F'(' '{print $1}' | sort -u
Author Custom Seccomp BPF Whitelist Profile in JSON
Create a least-privilege Seccomp profile setting defaultAction: SCMP_ACT_ERRNO (blocking everything by default) and explicitly allowing only essential syscalls (read, write, epoll_wait, futex, clone3, etc.).
{
"defaultAction": "SCMP_ACT_ERRNO",
"architectures": [
"SCMP_ARCH_X86_64",
"SCMP_ARCH_AARCH64"
],
"syscalls": [
{
"names": [
"read", "write", "openat", "close", "stat", "fstat",
"lseek", "mmap", "mprotect", "munmap", "brk", "rt_sigaction",
"rt_sigprocmask", "ioctl", "nanosleep", "epoll_create1",
"epoll_ctl", "epoll_wait", "clone3", "futex", "exit_group"
],
"action": "SCMP_ACT_ALLOW"
}
]
}
Apply and Validate Seccomp Profile on Running Container
Run the container referencing the custom seccomp profile with --security-opt seccomp=profile.json. Test that prohibited syscalls (e.g., sys_ptrace or sys_chroot) are instantly rejected with EPERM.
# Run container with custom seccomp profile
docker run -d \
--name hardened-api \
--security-opt seccomp=./strict-seccomp.json \
my-api:v1.0
# Test blocked syscall behavior inside the container
docker exec -it hardened-api strace -e ptrace ls
# Output: ptrace: Operation not permitted
Enforce AppArmor Mandatory Access Control (MAC)
Deploy an AppArmor profile in /etc/apparmor.d/docker-nginx-hardened to restrict file read/write access strictly to /etc/nginx and /usr/share/nginx/html, denying execution of /bin/sh or /usr/bin/curl.
# Load AppArmor profile into kernel
sudo apparmor_parser -r -W /etc/apparmor.d/docker-nginx-hardened
# Run container enforcing the AppArmor profile
docker run -d \
--security-opt apparmor=docker-nginx-hardened \
nginx:alpine
SCMP_ACT_ERRNO) alongside AppArmor MAC profiles confines compromised applications and renders kernel privilege escalation exploits ineffective."
- W
- e
- h
- a
- r
- d
- e
- n
- e
- d
- o
- u
- r
- i
- n
- t
- e
- r
- n
- e
- t
- -
- f
- a
- c
- i
- n
- g
- c
- o
- n
- t
- a
- i
- n
- e
- r
- t
- i
- e
- r
- a
- g
- a
- i
- n
- s
- t
- k
- e
- r
- n
- e
- l
- z
- e
- r
- o
- -
- d
- a
- y
- s
- b
- y
- a
- u
- d
- i
- t
- i
- n
- g
- r
- u
- n
- t
- i
- m
- e
- s
- y
- s
- c
- a
- l
- l
- s
- w
- i
- t
- h
- e
- B
- P
- F
- a
- n
- d
- g
- e
- n
- e
- r
- a
- t
- i
- n
- g
- s
- t
- r
- i
- c
- t
- S
- e
- c
- c
- o
- m
- p
- B
- P
- F
- w
- h
- i
- t
- e
- l
- i
- s
- t
- p
- r
- o
- f
- i
- l
- e
- s
- .
- P
- r
- o
- h
- i
- b
- i
- t
- i
- n
- g
- d
- a
- n
- g
- e
- r
- o
- u
- s
- s
- y
- s
- c
- a
- l
- l
- s
- l
- i
- k
- e
- `
- p
- t
- r
- a
- c
- e
- `
- a
- n
- d
- `
- b
- p
- f
- `
- c
- o
- m
- b
- i
- n
- e
- d
- w
- i
- t
- h
- A
- p
- p
- A
- r
- m
- o
- r
- f
- i
- l
- e
- s
- y
- s
- t
- e
- m
- c
- o
- n
- f
- i
- n
- e
- m
- e
- n
- t
- g
- u
- a
- r
- a
- n
- t
- e
- e
- s
- t
- h
- a
- t
- e
- v
- e
- n
- i
- f
- r
- e
- m
- o
- t
- e
- c
- o
- d
- e
- e
- x
- e
- c
- u
- t
- i
- o
- n
- o
- c
- c
- u
- r
- s
- ,
- t
- h
- e
- a
- d
- v
- e
- r
- s
- a
- r
- y
- c
- a
- n
- n
- o
- t
- t
- o
- u
- c
- h
- t
- h
- e
- h
- o
- s
- t
- k
- e
- r
- n
- e
- l
- o
- r
- e
- x
- e
- c
- u
- t
- e
- u
- n
- a
- u
- t
- h
- o
- r
- i
- z
- e
- d
- b
- i
- n
- a
- r
- i
- e
- s
- .
Senior hiring managers and bar raisers evaluate how you prioritize during outages. Avoid these common anti-patterns:
-
Structuring Dockerfile instructions inefficiently by copying app source before package manifests when configuring Hardening Containers Custom
Copying code before package manifests (
package.json,go.mod,requirements.txt) invalidates the build cache on every commit, forcing slow full reinstalls. -
Neglecting the PID 1 process reap problem and omitting proper SIGTERM signal forwarding when configuring Hardening Containers Custom
If an application runs as PID 1 without an init system (like
tini), it fails to reap zombie child processes and ignoresSIGTERMsignals, causing unclean 10s timeouts. - Assembling single-stage images containing compilers, build SDKs, and header files in production when configuring Hardening Containers Custom Single-stage builds bloat image size by gigabytes and expose extensive package vulnerabilities; multi-stage builds isolate artifacts into minimal distroless images.
- Confusing container port forwarding on bridge networks with unisolated host networking mode when configuring Hardening Containers Custom Bridge networking utilizes iptables NAT translations which add latency overhead, whereas host networking bypasses the network namespace entirely.