Merge pull request #1825 from oesteban/fix/DockerIssues

[FIX] Docker issues and documentation

Author: oesteban

CHANGES

@@ -1,6 +1,7 @@
 Upcoming release 0.13
 =====================
 
+* FIX: Issues in Docker image permissions, and docker documentation (https://github.com/nipy/nipype/pull/1825)
 * ENH: Revised all Dockerfiles and automated deployment to Docker Hub
        from CircleCI (https://github.com/nipy/nipype/pull/1815)
 * FIX: Semaphore capture using MultiProc plugin (https://github.com/nipy/nipype/pull/1689)

Dockerfile

@@ -182,7 +182,8 @@ RUN conda config --add channels conda-forge --add channels intel && \
                      libxslt=1.1.29 \
                      traits=4.6.0 \
                      psutil=5.0.1 \
-                     icu=58.1
+                     icu=58.1 && \
+    find /usr/local/miniconda/ -exec chmod 775 {} +
 
 # matplotlib cleanups: set default backend, precaching fonts
 RUN sed -i 's/\(backend *: \).*$/\1Agg/g' /usr/local/miniconda/lib/python3.5/site-packages/matplotlib/mpl-data/matplotlibrc && \

circle.yml

@@ -28,19 +28,17 @@ dependencies:
     - if [[ ! -d ~/examples/nipype-tutorial ]]; then wget --retry-connrefused --waitretry=5 --read-timeout=20 --timeout=15 -t 0 -q -O nipype-tutorial.tar.bz2 "${DATA_NIPYPE_TUTORIAL_URL}" && tar xjf nipype-tutorial.tar.bz2 -C ~/examples/; fi
     - if [[ ! -d ~/examples/nipype-fsl_course_data ]]; then wget --retry-connrefused --waitretry=5 --read-timeout=20 --timeout=15 -t 0 -q -O nipype-fsl_course_data.tar.gz "${DATA_NIPYPE_FSL_COURSE}" && tar xzf nipype-fsl_course_data.tar.gz -C ~/examples/; fi
     - if [[ ! -d ~/examples/feeds ]]; then wget --retry-connrefused --waitretry=5 --read-timeout=20 --timeout=15 -t 0 -q -O fsl-5.0.9-feeds.tar.gz "${DATA_NIPYPE_FSL_FEEDS}" && tar xzf fsl-5.0.9-feeds.tar.gz -C ~/examples/; fi
-    - if [[ -e $HOME/docker/image.tar ]]; then docker load -i $HOME/docker/image.tar; fi
-    - if [[ -e $HOME/docker/image27.tar ]]; then docker load -i $HOME/docker/image27.tar; fi
-    - if [[ -e $HOME/docker/image35.tar ]]; then docker load -i $HOME/docker/image35.tar; fi
+    - if [[ -e $HOME/docker/image.tar ]]; then docker load --input $HOME/docker/image.tar; else echo 'No docker image found in cache'; fi
+    - docker images
     - sed -i -E "s/(__version__ = )'[A-Za-z0-9.-]+'/\1'$CIRCLE_TAG'/" nipype/info.py
     - e=1 && for i in {1..5}; do docker build -t nipype/nipype:latest --build-arg BUILD_DATE=`date -u +"%Y-%m-%dT%H:%M:%SZ"` --build-arg VCS_REF=`git rev-parse --short HEAD` --build-arg VERSION=$CIRCLE_TAG . && e=0 && break || sleep 15; done && [ "$e" -eq "0" ] :
         timeout: 21600
-    - docker build -f docker/Dockerfile_py27 -t nipype/nipype_test:py27 . :
+    - e=1 && for i in {1..5}; do docker build -f docker/Dockerfile_py27 -t nipype/nipype_test:py27 . && e=0 && break || sleep 15; done && [ "$e" -eq "0" ] :
         timeout: 1600
-    - docker build -f docker/Dockerfile_py35 -t nipype/nipype_test:py35 . :
+    - e=1 && for i in {1..5}; do docker build -f docker/Dockerfile_py35 -t nipype/nipype_test:py35 . && e=0 && break || sleep 15; done && [ "$e" -eq "0" ] :
+        timeout: 1600
+    - docker save -o $HOME/docker/image.tar nipype/nipype:latest nipype/nipype_test:py27 nipype/nipype_test:py35 :
         timeout: 1600
-    - docker save nipype/nipype:latest > $HOME/docker/image.tar
-    - docker save nipype/nipype_test:py27 > $HOME/docker/image27.tar
-    - docker save nipype/nipype_test:py35 > $HOME/docker/image35.tar
 
 test:
   override:

doc/devel/testing_nipype.rst

@@ -107,26 +107,47 @@ This will skip any tests that require matlab.
 Testing Nipype using Docker
 ---------------------------
 
-As of :code:`nipype-0.13`, Nipype is tested inside Docker containers. Once the developer
-`has installed the Docker Engine <https://docs.docker.com/engine/installation/>`_, testing
-Nipype is as easy as follows::
+As of :code:`nipype-0.13`, Nipype is tested inside Docker containers. First, install the
+`Docker Engine <https://docs.docker.com/engine/installation/>`_.
+Nipype has one base docker image called nipype/nipype, and several additional test images
+for various Python versions.
+
+The base nipype/nipype image is built as follows::
 
   cd path/to/nipype/
-  docker build -f docker/nipype_test/Dockerfile_py27 -t nipype/nipype_test:py27
-  docker run -it --rm -v /etc/localtime:/etc/localtime:ro \
-                      -e FSL_COURSE_DATA="/root/examples/nipype-fsl_course_data" \
+  docker build -t nipype/nipype .
+
+This base image contains several useful tools (FreeSurfer, AFNI, FSL, ANTs, etc.) and
+a nipype installation, all in Python 3.5.
+It is possible to fetch a built image from the latest master branch of nipype
+using::
+
+  docker run -it --rm nipype/nipype:master
+
+
+The docker run command will then open the container and offer a bash shell for the
+developer.
+
+The additional test images have several test scripts installed. For instance,
+to build and run all tests on Python 2.7::
+
+  cd path/to/nipype/
+  docker build -f docker/Dockerfile_py27 -t nipype/nipype_test:py27 .
+  docker run -it --rm -e FSL_COURSE_DATA="/root/examples/nipype-fsl_course_data" \
                       -v ~/examples:/root/examples:ro \
                       -v ~/scratch:/scratch \
                       -w /root/src/nipype \
-                      nipype/nipype_test:py27 /usr/bin/run_pytest.sh
+                      nipype/nipype_test:py27 /usr/bin/run_pytests.sh
 
 For running nipype in Python 3.5::
 
   cd path/to/nipype/
-  docker build -f docker/nipype_test/Dockerfile_py35 -t nipype/nipype_test:py35
-  docker run -it --rm -v /etc/localtime:/etc/localtime:ro \
-                      -e FSL_COURSE_DATA="/root/examples/nipype-fsl_course_data" \
+  docker build -f docker/Dockerfile_py35 -t nipype/nipype_test:py35 .
+  docker run -it --rm -e FSL_COURSE_DATA="/root/examples/nipype-fsl_course_data" \
                       -v ~/examples:/root/examples:ro \
                       -v ~/scratch:/scratch \
                       -w /root/src/nipype \
-                      nipype/nipype_test:py35 /usr/bin/run_pytest.sh
+                      nipype/nipype_test:py35 /usr/bin/run_pytests.sh
+
+The last two examples assume that the example data is downladed into ~/examples and
+the ~/scratch folder will be created if it does not exist previously.