docker-library / docker-library/python

Debbuging a `-slim` image?

Abierto
#903 1 comentario 0 reacciones 0 asignados Ver en GitHub

Nadie ha tomado este issue todavía.

Lenguaje dominante
Dockerfile
Estrellas
2.8k
Forks
1.1k
Merge medio
3 d 1 h
PR fusionados (30 d)
1

Descripción

We are hitting a rare deadlock in production that can't be reproduced using debug images.
Only way to debug it is to attach to the production image as the problem happens.
Problem is that production are based on the -slim images.

apt install gdb and python3-dbg is not enough because the python running is built from source and not aligned with what is available in apt.

I've tried to start the corresponding non-slim image but their binaries don't seem to align enough to make gdb happy:

Start python inside -slim image

➡️ My end goal is to run gdb py-bt on this process:
docker run -ti python:3.11-slim-bookworm python3 -c "import time; time.sleep(1000)"

Inside this image there is no chance to debug at all due to missing gdb and debug symbols. This is expected.

Build a debugger image from the corresponding non-slim image.
FROM python:3.11-bookworm

RUN --mount=type=cache,target=/var/cache/apt --mount=type=cache,target=/var/lib/apt \
    apt update -y && \
    apt install -y gdb

docker build -t mydebugpython .

Use the newly built debug image to attach into the first running container

Find the container id and use that to attach to the same pid-namespace.

docker run \
    --cap-add SYS_PTRACE \
    --pid container:15aab7ea5f57 \
    --privileged  \
    -ti \
    --entrypoint bash docker.io/library/mydebugpython
root@bb90d04f8722:/# gdb -q python --pid 1 -ex "py-bt"

Reading symbols from python...
Attaching to program: /usr/local/bin/python, process 1

warning: Build ID mismatch between current exec-file /usr/local/bin/python
and automatically determined exec-file /usr/local/bin/python3.11
exec-file-mismatch handling is currently "ask"

Load new symbol table from "/usr/local/bin/python3.11"? (y or n) y
Reading symbols from /usr/local/bin/python3.11...
Reading symbols from target:/usr/local/bin/../lib/libpython3.11.so.1.0...
(No debugging symbols found in target:/usr/local/bin/../lib/libpython3.11.so.1.0)
Reading symbols from target:/lib/x86_64-linux-gnu/libc.so.6...
Reading symbols from /usr/lib/debug/.build-id/82/ce4e6e4ef08fa58a3535f7437bd3e592db5ac0.debug...
Reading symbols from target:/lib/x86_64-linux-gnu/libm.so.6...
Reading symbols from /usr/lib/debug/.build-id/ea/87e1b3daf095cd53f1f99ab34a88827eccce80.debug...
Reading symbols from target:/lib64/ld-linux-x86-64.so.2...
Reading symbols from /usr/lib/debug/.build-id/38/e7d4a67acf053c794b3b8094e6900b5163f37d.debug...
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
__GI___clock_nanosleep (clock_id=1, flags=1, req=0x7ffd9d9ec458, rem=0x0) at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:71
71	../sysdeps/unix/sysv/linux/clock_nanosleep.c: No such file or directory.
Traceback (most recent call first):
  (unable to read python frame information)    # <--------------   :(

So it seems like the difference of slim and non-slim isn't only the presence of debug symbols or not, but also the binary is built differently?
Is this possible to solve somehow? My knowledge of how python is built and how to operate gdb are only moderate.

Guía de contribución

No hay ninguna guía de contribución indexada para este repositorio

Primeros pasos

  1. Lee el issue completo y luego la guía de contribución del proyecto.
  2. Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
  3. Haz un fork del repositorio y trabaja en una rama.
  4. Abre un pull request que haga referencia al número del issue.

Línea de trabajo

No se ha indicado ningún archivo del repositorio ni ninguna prueba. Empieza comparando las compilaciones de imagen python:3.11-slim-bookworm y python:3.11-bookworm, luego reproduce los comandos de attach de gdb e inspecciona la discrepancia entre el ejecutable y los símbolos de depuración. Se considera terminado cuando se haya establecido si py-bt puede funcionar con la imagen slim y se haya documentado un enfoque de depuración reproducible o una limitación.

Escrito por el modelo de indexación a partir del texto del issue.

Evaluación

Stack tecnológico
docker, python
Área
devops, tooling
Tipo de issue
Error
Dificultad
4/5
Tiempo estimado
3-5 días
Estado de actividad
Estancado
Claridad
Bastante claro
Aptitud para principiantes
25/100

Recibe los nuevos issues en tu correo

Un resumen breve de issues de GitHub para principiantes.