ansible/library/cloud/s3

458 lines
18 KiB
Text
Raw Normal View History

#!/usr/bin/python
2013-03-16 12:18:43 +01:00
# This file is part of Ansible
#
# Ansible is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Ansible is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Ansible. If not, see <http://www.gnu.org/licenses/>.
DOCUMENTATION = '''
---
module: s3
short_description: idempotent S3 module putting a file into S3.
2013-03-16 12:18:43 +01:00
description:
- This module allows the user to dictate the presence of a given file in an S3 bucket. If or once the key (file) exists in the bucket, it returns a time-expired download URL. This module has a dependency on python-boto.
2013-03-16 12:18:43 +01:00
version_added: "1.1"
options:
bucket:
description:
- Bucket name.
2013-03-16 12:18:43 +01:00
required: true
default: null
aliases: []
object:
2013-03-16 12:18:43 +01:00
description:
- Keyname of the object inside the bucket. Can be used to create "virtual directories", see examples.
2013-03-16 12:18:43 +01:00
required: false
default: null
aliases: []
version_added: "1.3"
src:
description:
- The source file path when performing a PUT operation.
required: false
default: null
aliases: []
version_added: "1.3"
dest:
2013-03-16 12:18:43 +01:00
description:
- The destination file path when downloading an object/key with a GET operation.
2013-03-16 12:18:43 +01:00
required: false
aliases: []
version_added: "1.3"
overwrite:
description:
- Force overwrite either locally on the filesystem or remotely with the object/key. Used with PUT and GET operations.
required: false
default: true
2013-04-05 21:18:31 +02:00
version_added: "1.2"
mode:
description:
- Switches the module behaviour between put (upload), get (download), geturl (return download url (Ansible 1.3+), getstr (download object as string (1.3+)), create (bucket) and delete (bucket).
required: true
default: null
aliases: []
expiration:
description:
- Time limit (in seconds) for the URL generated and returned by S3/Walrus when performing a mode=put or mode=geturl operation.
required: false
default: 600
aliases: []
s3_url:
description:
- S3 URL endpoint. If not specified then the S3_URL environment variable is used, if that variable is defined.
default: null
aliases: [ S3_URL ]
aws_secret_key:
description:
- AWS secret key. If not set then the value of the AWS_SECRET_KEY environment variable is used.
required: false
default: null
aliases: ['ec2_secret_key', 'secret_key']
aws_access_key:
description:
- AWS access key. If not set then the value of the AWS_ACCESS_KEY environment variable is used.
required: false
default: null
aliases: [ 'ec2_access_key', 'access_key' ]
2013-03-16 12:18:43 +01:00
requirements: [ "boto" ]
author: Lester Wade, Ralph Tice
2013-03-16 12:18:43 +01:00
'''
EXAMPLES = '''
# Simple PUT operation
- s3: bucket=mybucket object=/my/desired/key.txt src=/usr/local/myfile.txt mode=put
# Simple GET operation
- s3: bucket=mybucket object=/my/desired/key.txt dest=/usr/local/myfile.txt mode=get
# GET/download and overwrite local file (trust remote)
- s3: bucket=mybucket object=/my/desired/key.txt dest=/usr/local/myfile.txt mode=get
# GET/download and do not overwrite local file (trust remote)
- s3: bucket=mybucket object=/my/desired/key.txt dest=/usr/local/myfile.txt mode=get force=false
# PUT/upload and overwrite remote file (trust local)
- s3: bucket=mybucket object=/my/desired/key.txt src=/usr/local/myfile.txt mode=put
# PUT/upload and do not overwrite remote file (trust local)
- s3: bucket=mybucket object=/my/desired/key.txt src=/usr/local/myfile.txt mode=put force=false
# Download an object as a string to use else where in your playbook
- s3: bucket=mybucket object=/my/desired/key.txt src=/usr/local/myfile.txt mode=getstr
# Create an empty bucket
- s3: bucket=mybucket mode=create
# Create a bucket with key as directory
- s3: bucket=mybucket object=/my/directory/path mode=create
# Delete a bucket and all contents
- s3: bucket=mybucket mode=delete
'''
2013-03-16 12:18:43 +01:00
import sys
import os
import urlparse
import hashlib
2013-03-16 12:18:43 +01:00
try:
import boto
except ImportError:
module.fail_json(msg="boto required for this module")
2013-03-16 12:18:43 +01:00
def key_check(module, s3, bucket, obj):
try:
bucket = s3.lookup(bucket)
key_check = bucket.get_key(obj)
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
if key_check:
return True
else:
return False
def keysum(module, s3, bucket, obj):
bucket = s3.lookup(bucket)
key_check = bucket.get_key(obj)
if key_check:
md5_remote = key_check.etag[1:-1]
etag_multipart = md5_remote.find('-')!=-1 #Check for multipart, etag is not md5
if etag_multipart is True:
module.fail_json(msg="Files uploaded with multipart of s3 are not supported with checksum, unable to compute checksum.")
return md5_remote
def bucket_check(module, s3, bucket):
try:
result = s3.lookup(bucket)
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
if result:
return True
else:
return False
def create_bucket(module, s3, bucket):
try:
bucket = s3.create_bucket(bucket)
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
if bucket:
return True
def delete_bucket(module, s3, bucket):
try:
bucket = s3.lookup(bucket)
bucket_contents = bucket.list()
bucket.delete_keys([key.name for key in bucket_contents])
bucket.delete()
return True
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
def delete_key(module, s3, bucket, obj):
try:
bucket = s3.lookup(bucket)
bucket.delete_key(obj)
module.exit_json(msg="Object deleted from bucket %s"%bucket, changed=True)
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
def create_dirkey(module, s3, bucket, obj):
try:
bucket = s3.lookup(bucket)
key = bucket.new_key(obj)
key.set_contents_from_string('')
module.exit_json(msg="Virtual directory %s created in bucket %s" % (obj, bucket.name), changed=True)
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
def upload_file_check(src):
if os.path.exists(src):
file_exists is True
else:
file_exists is False
if os.path.isdir(src):
module.fail_json(msg="Specifying a directory is not a valid source for upload.", failed=True)
return file_exists
def path_check(path):
if os.path.exists(path):
return True
else:
return False
def upload_s3file(module, s3, bucket, obj, src, expiry):
try:
bucket = s3.lookup(bucket)
key = bucket.new_key(obj)
key.set_contents_from_filename(src)
url = key.generate_url(expiry)
module.exit_json(msg="PUT operation complete", url=url, changed=True)
except s3.provider.storage_copy_error, e:
module.fail_json(msg= str(e))
def download_s3file(module, s3, bucket, obj, dest):
try:
bucket = s3.lookup(bucket)
key = bucket.lookup(obj)
key.get_contents_to_filename(dest)
module.exit_json(msg="GET operation complete", changed=True)
except s3.provider.storage_copy_error, e:
module.fail_json(msg= str(e))
def download_s3str(module, s3, bucket, obj):
try:
bucket = s3.lookup(bucket)
key = bucket.lookup(obj)
contents = key.get_contents_as_string()
module.exit_json(msg="GET operation complete", contents=contents, changed=True)
except s3.provider.storage_copy_error, e:
module.fail_json(msg= str(e))
def get_download_url(module, s3, bucket, obj, expiry, changed=True):
try:
bucket = s3.lookup(bucket)
key = bucket.lookup(obj)
url = key.generate_url(expiry)
module.exit_json(msg="Download url:", url=url, expiry=expiry, changed=changed)
except s3.provider.storage_response_error, e:
module.fail_json(msg= str(e))
def is_walrus(s3_url):
""" Return True if it's Walrus endpoint, not S3
We assume anything other than *.amazonaws.com is Walrus"""
if s3_url is not None:
o = urlparse.urlparse(s3_url)
return not o.hostname.endswith('amazonaws.com')
else:
return False
2013-03-16 12:18:43 +01:00
def main():
argument_spec = ec2_argument_keys_spec()
argument_spec.update(dict(
bucket = dict(required=True),
object = dict(),
src = dict(),
2013-10-02 09:57:54 +02:00
dest = dict(default=None),
mode = dict(choices=['get', 'put', 'delete', 'create', 'geturl', 'getstr'], required=True),
2013-06-01 18:09:04 +02:00
expiry = dict(default=600, aliases=['expiration']),
s3_url = dict(aliases=['S3_URL']),
overwrite = dict(aliases=['force'], default=True, type='bool'),
)
2013-03-16 12:18:43 +01:00
)
module = AnsibleModule(argument_spec=argument_spec)
2013-03-16 12:18:43 +01:00
bucket = module.params.get('bucket')
obj = module.params.get('object')
src = module.params.get('src')
2013-10-02 09:57:54 +02:00
if module.params.get('dest'):
dest = os.path.expanduser(module.params.get('dest'))
mode = module.params.get('mode')
2013-03-16 12:18:43 +01:00
expiry = int(module.params['expiry'])
s3_url = module.params.get('s3_url')
2013-06-01 18:09:04 +02:00
overwrite = module.params.get('overwrite')
ec2_url, aws_access_key, aws_secret_key, region = get_ec2_creds(module)
if module.params.get('object'):
obj = os.path.expanduser(module.params['object'])
2013-03-16 12:18:43 +01:00
# allow eucarc environment variables to be used if ansible vars aren't set
if not s3_url and 'S3_URL' in os.environ:
s3_url = os.environ['S3_URL']
2013-03-16 12:18:43 +01:00
# If we have an S3_URL env var set, this is likely to be Walrus, so change connection method
if is_walrus(s3_url):
2013-03-16 12:18:43 +01:00
try:
walrus = urlparse.urlparse(s3_url).hostname
s3 = boto.connect_walrus(walrus, aws_access_key, aws_secret_key)
2013-03-16 12:18:43 +01:00
except boto.exception.NoAuthHandlerFound, e:
module.fail_json(msg = str(e))
else:
try:
s3 = boto.connect_s3(aws_access_key, aws_secret_key)
2013-03-16 12:18:43 +01:00
except boto.exception.NoAuthHandlerFound, e:
module.fail_json(msg = str(e))
# If our mode is a GET operation (download), go through the procedure as appropriate ...
if mode == 'get':
2013-03-16 12:18:43 +01:00
# First, we check to see if the bucket exists, we get "bucket" returned.
bucketrtn = bucket_check(module, s3, bucket)
if bucketrtn is False:
module.fail_json(msg="Target bucket cannot be found", failed=True)
2013-03-16 12:18:43 +01:00
# Next, we check to see if the key in the bucket exists. If it exists, it also returns key_matches md5sum check.
keyrtn = key_check(module, s3, bucket, obj)
if keyrtn is False:
module.fail_json(msg="Target key cannot be found", failed=True)
2013-03-16 12:18:43 +01:00
# If the destination path doesn't exist, no need to md5um etag check, so just download.
pathrtn = path_check(dest)
if pathrtn is False:
download_s3file(module, s3, bucket, obj, dest)
# Compare the remote MD5 sum of the object with the local dest md5sum, if it already exists.
if pathrtn is True:
md5_remote = keysum(module, s3, bucket, obj)
md5_local = hashlib.md5(open(dest, 'rb').read()).hexdigest()
if md5_local == md5_remote:
sum_matches = True
if overwrite is True:
download_s3file(module, s3, bucket, obj, dest)
else:
module.exit_json(msg="Local and remote object are identical, ignoring. Use overwrite parameter to force.", changed=False)
2013-03-16 12:18:43 +01:00
else:
sum_matches = False
if overwrite is True:
download_s3file(module, s3, bucket, obj, dest)
else:
module.fail_json(msg="WARNING: Checksums do not match. Use overwrite parameter to force download.", failed=True)
# Firstly, if key_matches is TRUE and overwrite is not enabled, we EXIT with a helpful message.
if sum_matches is True and overwrite is False:
module.exit_json(msg="Local and remote object are identical, ignoring. Use overwrite parameter to force.", changed=False)
# At this point explicitly define the overwrite condition.
if sum_matches is True and pathrtn is True and overwrite is True:
download_s3file(module, s3, bucket, obj, dest)
# If sum does not match but the destination exists, we
# if our mode is a PUT operation (upload), go through the procedure as appropriate ...
if mode == 'put':
# Use this snippet to debug through conditionals:
# module.exit_json(msg="Bucket return %s"%bucketrtn)
# sys.exit(0)
# Lets check the src path.
pathrtn = path_check(src)
if pathrtn is False:
module.fail_json(msg="Local object for PUT does not exist", failed=True)
# Lets check to see if bucket exists to get ground truth.
bucketrtn = bucket_check(module, s3, bucket)
if bucketrtn is True:
keyrtn = key_check(module, s3, bucket, obj)
# Lets check key state. Does it exist and if it does, compute the etag md5sum.
if bucketrtn is True and keyrtn is True:
md5_remote = keysum(module, s3, bucket, obj)
md5_local = hashlib.md5(open(src, 'rb').read()).hexdigest()
if md5_local == md5_remote:
sum_matches = True
if overwrite is True:
upload_s3file(module, s3, bucket, obj, src, expiry)
else:
2013-10-09 18:25:09 +02:00
get_download_url(module, s3, bucket, obj, expiry, changed=False)
else:
sum_matches = False
if overwrite is True:
upload_s3file(module, s3, bucket, obj, src, expiry)
else:
module.exit_json(msg="WARNING: Checksums do not match. Use overwrite parameter to force upload.", failed=True)
# If neither exist (based on bucket existence), we can create both.
if bucketrtn is False and pathrtn is True:
create_bucket(module, s3, bucket)
upload_s3file(module, s3, bucket, obj, src, expiry)
2013-03-16 12:18:43 +01:00
# If bucket exists but key doesn't, just upload.
if bucketrtn is True and pathrtn is True and keyrtn is False:
upload_s3file(module, s3, bucket, obj, src, expiry)
2013-03-16 12:18:43 +01:00
# Support for deleting an object if we have both params.
if mode == 'delete':
if bucket:
bucketrtn = bucket_check(module, s3, bucket)
if bucketrtn is True:
deletertn = delete_bucket(module, s3, bucket)
if deletertn is True:
module.exit_json(msg="Bucket %s and all keys have been deleted."%bucket, changed=True)
else:
2013-10-07 14:59:03 +02:00
module.fail_json(msg="Bucket does not exist.", changed=False)
2013-03-16 12:18:43 +01:00
else:
module.fail_json(msg="Bucket parameter is required.", failed=True)
# Need to research how to create directories without "populating" a key, so this should just do bucket creation for now.
# WE SHOULD ENABLE SOME WAY OF CREATING AN EMPTY KEY TO CREATE "DIRECTORY" STRUCTURE, AWS CONSOLE DOES THIS.
if mode == 'create':
if bucket and not obj:
bucketrtn = bucket_check(module, s3, bucket)
if bucketrtn is True:
module.exit_json(msg="Bucket already exists.", changed=False)
else:
2013-10-02 09:57:54 +02:00
module.exit_json(msg="Bucket created succesfully", changed=create_bucket(module, s3, bucket))
if bucket and obj:
bucketrtn = bucket_check(module, s3, bucket)
if obj.endswith('/'):
dirobj = obj
else:
dirobj = obj + "/"
if bucketrtn is True:
keyrtn = key_check(module, s3, bucket, dirobj)
if keyrtn is True:
module.exit_json(msg="Bucket %s and key %s already exists."% (bucket, obj), changed=False)
else:
create_dirkey(module, s3, bucket, dirobj)
if bucketrtn is False:
created = create_bucket(module, s3, bucket)
create_dirkey(module, s3, bucket, dirobj)
# Support for grabbing the time-expired URL for an object in S3/Walrus.
if mode == 'geturl':
if bucket and obj:
bucketrtn = bucket_check(module, s3, bucket)
if bucketrtn is False:
module.fail_json(msg="Bucket %s does not exist."%bucket, failed=True)
else:
keyrtn = key_check(module, s3, bucket, obj)
if keyrtn is True:
get_download_url(module, s3, bucket, obj, expiry)
else:
module.fail_json(msg="Key %s does not exist."%obj, failed=True)
else:
module.fail_json(msg="Bucket and Object parameters must be set", failed=True)
if mode == 'getstr':
if bucket and obj:
bucketrtn = bucket_check(module, s3, bucket)
if bucketrtn is False:
module.fail_json(msg="Bucket %s does not exist."%bucket, failed=True)
else:
keyrtn = key_check(module, s3, bucket, obj)
if keyrtn is True:
download_s3str(module, s3, bucket, obj)
else:
module.fail_json(msg="Key %s does not exist."%obj, failed=True)
module.exit_json(failed=False)
# import module snippets
from ansible.module_utils.basic import *
from ansible.module_utils.ec2 import *
2013-03-16 12:18:43 +01:00
main()