Merge lp:~harlowja/cloud-init/query-back-duo into lp:~cloud-init-dev/cloud-init/trunk

[Joshua Harlow]

Reworking the query tool back to life addition. This method is a little less intrusive.

[Server Team CI bot (server-team-bot)]

FAILED: Continuous integration, rev:755
https://server-team-jenkins.canonical.com/job/cloud-init-ci/58/
Executed test runs:
    None: https://server-team-jenkins.canonical.com/job/lp-vote-on-merge/32/console

Click here to trigger a rebuild:
https://server-team-jenkins.canonical.com/job/cloud-init-ci/58/rebuild

review: Needs Fixing (continuous-integration)

[Chad Smith (chad.smith)]

Hello,
Thank you for taking the time to contribute to cloud-init. Cloud-init has moved its revision control system to git. We are cleaning up old reviews that no longer apply to the current codebase. As a result, we are marking all bzr merge proposals as 'rejected'. If you would like to re-submit this proposal for review, please do so by following the current HACKING documentation at http://cloudinit.readthedocs.io/en/latest/topics/hacking.html.

Also, as you know= cloud-init cmdline util has now moved to cloudinit/cmd/main.py

review: Needs Fixing

Unmerged revisions

755. By Joshua Harlow on 2012-12-24

Integrate a slightly cleaner query tool, bring it back to life!

=== modified file 'bin/cloud-init'
--- bin/cloud-init	2012-09-20 01:19:43 +0000
+++ bin/cloud-init	2012-12-24 04:32:20 +0000
@@ -26,6 +26,8 @@
 import sys
 import traceback
 
+import cPickle as pickle
+
 # This is more just for running from the bin folder so that
 # cloud-init binary can find the cloudinit module
 possible_topdir = os.path.normpath(os.path.join(os.path.abspath(
@@ -45,6 +47,8 @@
 from cloudinit import util
 from cloudinit import version
 
+from cloudinit.sources import DataSource
+
 from cloudinit.settings import (PER_INSTANCE, PER_ALWAYS, PER_ONCE,
                                 CLOUD_CONFIG)
 
@@ -56,12 +60,18 @@
 # Module section template
 MOD_SECTION_TPL = "cloud_%s_modules"
 
-# Things u can query on
-QUERY_DATA_TYPES = [
-    'data',
-    'data_raw',
-    'instance_id',
-]
+# Things u can query on and the lambda to fetch that
+# item when given a datasource to apply on
+QUERY_DATA_TYPES = {
+    'metadata': lambda ds: ds.metadata,
+    'userdata': lambda ds: ds.userdata,
+    'iid': lambda ds: ds.get_instance_id(),
+    'hostname': lambda ds: ds.get_hostname(),
+    'locale': lambda ds: ds.get_locale(),
+    'az': lambda ds: ds.availability_zone,
+    'launch_index': lambda ds: ds.launch_index,
+    'public_keys': lambda ds: ds.get_public_ssh_keys(),
+}
 
 # Frequency shortname to full name
 # (so users don't have to remember the full name...)
@@ -86,6 +96,10 @@
     sys.stderr.write("\n")
 
 
+def is_root():
+    return os.geteuid() == 0
+
+
 def welcome(action, msg=None):
     if not msg:
         msg = welcome_format(action)
@@ -346,9 +360,48 @@
     return run_module_section(mods, name, name)
 
 
-def main_query(name, _args):
-    raise NotImplementedError(("Action '%s' is not"
-                               " currently implemented") % (name))
+def main_query(name, args):
+    # TODO(harlowja) refactor init so that
+    # it is not the sole entrypoint for so many operations
+    # perhaps split into
+    init = stages.Init(ds_deps=[])
+    safe_ds_where = init.paths.get_ipath_cur("obj_pkl_safe")
+    if not os.path.isfile(safe_ds_where):
+        if is_root() and args.populate:
+            # At this point we can make a safe datasource from whatever
+            # may exist locally (if anything)
+            ds = init.fetch(True)
+            if not ds:
+                raise NotImplementedError(("No datasource found to "
+                                           "convert into a 'safe' datasource"))
+            else:
+                # Keep only the basics, ignoring the userdata since typically
+                # said userdata contains contents and data that users should
+                # not always be able to see.
+                safe_ds = DataSource(ds.sys_cfg, ds.distro, ds.paths)
+                safe_ds.metadata = ds.metadata
+                pk_contents = pickle.dumps(safe_ds)
+                util.write_file(safe_ds_where, pk_contents, mode=0644)
+    if args.what:
+        ds = None
+        if is_root():
+            # Use a full datasource if running as root
+            ds = init.fetch(True)
+        if not ds:
+            try:
+                pk_contents = util.load_file(safe_ds_where)
+                ds = pickle.loads(pk_contents)
+            except:
+                pass
+        if not ds:
+            raise NotImplementedError("No datasource found for querying.")
+        for i in args.what:
+            i_canon = i.lower().strip()
+            if i_canon not in QUERY_DATA_TYPES:
+                print("Unknown how to query on %s!" % (i))
+            else:
+                func = QUERY_DATA_TYPES[i_canon]
+                print("%s: %s" % (i, func(ds)))
 
 
 def main_single(name, args):
@@ -468,10 +521,14 @@
     parser_query = subparsers.add_parser('query',
                                          help=('query information stored '
                                                'in cloud-init'))
-    parser_query.add_argument("--name", '-n', action="store",
-                              help="item name to query on",
-                              required=True,
-                              choices=QUERY_DATA_TYPES)
+    if is_root():
+        parser_query.add_argument("--populate", '-p', action="store_true",
+                                  help=("populate the 'safe' "
+                                        "datasource that users can query"),
+                                  default=False)
+    query_help = "query the given datasource field (%s)"
+    query_help = query_help % (", ".join(sorted(QUERY_DATA_TYPES.keys())))
+    parser_query.add_argument('what', nargs='*', help=query_help)
     parser_query.set_defaults(action=('query', main_query))
 
     # This subcommand allows you to run a single module

=== modified file 'cloudinit/sources/__init__.py'
--- cloudinit/sources/__init__.py	2012-11-12 17:26:49 +0000
+++ cloudinit/sources/__init__.py	2012-12-24 04:32:20 +0000
@@ -20,7 +20,6 @@
 #    You should have received a copy of the GNU General Public License
 #    along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
-import abc
 import os
 
 from cloudinit import importer
@@ -42,9 +41,6 @@
 
 
 class DataSource(object):
-
-    __metaclass__ = abc.ABCMeta
-
     def __init__(self, sys_cfg, distro, paths, ud_proc=None):
         self.sys_cfg = sys_cfg
         self.distro = distro

=== modified file 'cloudinit/stages.py'
--- cloudinit/stages.py	2012-12-17 13:41:11 +0000
+++ cloudinit/stages.py	2012-12-24 04:32:20 +0000
@@ -217,13 +217,14 @@
         cfg_list = self.cfg.get('datasource_list') or []
         return (cfg_list, pkg_list)
 
-    def _get_data_source(self):
+    def _get_data_source(self, local_only):
         if self.datasource is not NULL_DATA_SOURCE:
             return self.datasource
         ds = self._restore_from_cache()
         if ds:
             LOG.debug("Restored from cache, datasource: %s", ds)
-        if not ds:
+        # Try to find one that will work
+        if not ds and not local_only:
             (cfg_list, pkg_list) = self._get_datasources()
             # Deep copy so that user-data handlers can not modify
             # (which will affect user-data handlers down the line...)
@@ -303,8 +304,8 @@
         self._reset()
         return iid
 
-    def fetch(self):
-        return self._get_data_source()
+    def fetch(self, local_only=False):
+        return self._get_data_source(local_only)
 
     def instancify(self):
         return self._reflect_cur_instance()