Relocate xml tasks and switch to using bz2.

playbooks/roles/edxapp/tasks/deploy.yml

@@ -219,37 +219,6 @@
   - "restart edxapp"
   - "restart edxapp_workers"
 
-# Need to do this because s3 module fails when trying
-# to checksum a multipart s3 object. And overwrite is
-# not effective
-- name: remove existing coursedata file
-  file: path="/var/tmp/xml_coursedata.tar.gz" state=absent
-
-# Install the xml courses from an s3 bucket
-- name: download xml coursedata from s3
-  s3: >
-    bucket="{{ EDXAPP_XML_S3_BUCKET }}"
-    object="{{ EDXAPP_XML_S3_KEY }}"
-    mode="get"
-    dest="/var/tmp/xml_coursedata.tar.gz"
-  when: not EDXAPP_XML_FROM_GIT and EDXAPP_XML_S3_BUCKET and EDXAPP_XML_S3_KEY
-  register: download_xml_s3
-
-# Install the xml courses from a url.
-- name: download xml coursedata
-  get_url: url="{{ EDXAPP_XML_URL }}" dest="/var/tmp/xml_coursedata.tar.gz"
-  when: not EDXAPP_XML_FROM_GIT and EDXAPP_XML_URL
-  register: download_xml_url
-
-- name: unzip the data to the data dir
-  shell: >
-    tar xvzf /var/tmp/xml_coursedata.tar.gz
-    chdir="{{ edxapp_course_data_dir }}"
-  when: download_xml_s3.changed or download_xml_url.changed
-
-- include: xml.yml
-  when: EDXAPP_XML_FROM_GIT
-
 # If using CAS and you have a function for mapping attributes, install
 # the module here.  The next few tasks set up the python code sandbox
 - name: install CAS attribute module
@@ -280,6 +249,39 @@
   - "restart edxapp"
   - "restart edxapp_workers"
 
+# The next few tasks install xml courses.
+# Need to do this because s3 module fails when trying
+# to checksum a multipart s3 object. And overwrite is
+# not effective
+- name: remove existing coursedata file
+  file: path="/var/tmp/xml_coursedata.tar.gz" state=absent
+  when: not EDXAPP_XML_FROM_GIT
+
+# Install the xml courses from an s3 bucket
+- name: download xml coursedata from s3
+  s3: >
+    bucket="{{ EDXAPP_XML_S3_BUCKET }}"
+    object="{{ EDXAPP_XML_S3_KEY }}"
+    mode="get"
+    dest="/var/tmp/xml_coursedata.tar.gz"
+  when: not EDXAPP_XML_FROM_GIT and EDXAPP_XML_S3_BUCKET and EDXAPP_XML_S3_KEY
+  register: download_xml_s3
+
+# Install the xml courses from a url.
+- name: download xml coursedata
+  get_url: url="{{ EDXAPP_XML_URL }}" dest="/var/tmp/xml_coursedata.tar.bz2"
+  when: not EDXAPP_XML_FROM_GIT and EDXAPP_XML_URL
+  register: download_xml_url
+
+- name: unzip the data to the data dir
+  shell: >
+    tar xvjf /var/tmp/xml_coursedata.tar.bz2
+    chdir="{{ edxapp_course_data_dir }}"
+  when: download_xml_s3.changed or download_xml_url.changed
+
+- include: xml.yml
+  when: EDXAPP_XML_FROM_GIT
+
 # The next few tasks set up the python code sandbox
 
 # need to disable this profile, otherwise the pip inside the sandbox venv has no permissions

playbooks/roles/edxapp/tasks/xml.yml

@@ -50,7 +50,7 @@
   when: item.disposition == "import"
 
 - name: create an archive of course data and course static dirs
-  shell: tar czf /tmp/static_course_content.tar.gz -C {{ edxapp_data_dir }} {{ edxapp_course_data_dir|basename }} {{ edxapp_course_static_dir|basename }}
+  shell: tar cjf /tmp/static_course_content.tar.bz2 -C {{ edxapp_data_dir }} {{ edxapp_course_data_dir|basename }} {{ edxapp_course_static_dir|basename }}
   when: EDXAPP_XML_S3_BUCKET and EDXAPP_XML_S3_KEY
 
 - name: upload archive to s3
@@ -59,9 +59,9 @@
     object="{{ EDXAPP_XML_S3_KEY }}"
     mode=put
     overwrite=True
-    src="/tmp/static_course_content.tar.gz"
+    src="/tmp/static_course_content.tar.bz2"
   when: EDXAPP_XML_S3_BUCKET and EDXAPP_XML_S3_KEY
 
 - name: remove archive from disk
-  file: path="/tmp/static_course_content.tar.gz" state=absent
+  file: path="/tmp/static_course_content.tar.bz2" state=absent
   when: EDXAPP_XML_S3_BUCKET and EDXAPP_XML_S3_KEY