-# Copyright 2011-2013 GRNET S.A. All rights reserved.
+# Copyright 2011-2014 GRNET S.A. All rights reserved.
#
# Redistribution and use in source and binary forms, with or
# without modification, are permitted provided that the following
from kamaki.clients import SilentEvent, sendlog
from kamaki.clients.pithos.rest_api import PithosRestClient
from kamaki.clients.storage import ClientError
-from kamaki.clients.utils import path4url, filter_in
+from kamaki.clients.utils import path4url, filter_in, readall
def _pithos_hash(block, blockhash):
self.container = cnt_back_up
def purge_container(self, container=None):
- """Delete an empty container and destroy associated blocks
- """
+ """Delete an empty container and destroy associated blocks"""
cnt_back_up = self.container
try:
self.container = container or cnt_back_up
raise ClientError(msg, 1)
f = StringIO(data)
else:
- data = f.read(size) if size else f.read()
+ data = readall(f, size) if size else f.read()
r = self.object_put(
obj,
data=data,
hash_gen = hash_cb(nblocks)
hash_gen.next()
- for i in range(nblocks):
- block = fileobj.read(min(blocksize, size - offset))
+ for i in xrange(nblocks):
+ block = readall(fileobj, min(blocksize, size - offset))
bytes = len(block)
+ if bytes <= 0:
+ break
hash = _pithos_hash(block, blockhash)
hashes.append(hash)
hmap[hash] = (offset, bytes)
offset += bytes
if hash_cb:
hash_gen.next()
- msg = 'Failed to calculate uploaded blocks:'
- ' Offset and object size do not match'
+ msg = ('Failed to calculate uploading blocks: '
+ 'read bytes(%s) != requested size (%s)' % (offset, size))
assert offset == size, msg
def _upload_missing_blocks(self, missing, hmap, fileobj, upload_gen=None):
for hash in missing:
offset, bytes = hmap[hash]
fileobj.seek(offset)
- data = fileobj.read(bytes)
+ data = readall(fileobj, bytes)
r = self._put_block_async(data, hash)
flying.append(r)
unfinished = self._watch_thread_limit(flying)
blocksize, blockhash, size, nblocks) = self._get_file_block_info(
f, size, container_info_cache)
(hashes, hmap, offset) = ([], {}, 0)
- if not content_type:
- content_type = 'application/octet-stream'
+ content_type = content_type or 'application/octet-stream'
self._calculate_blocks_for_upload(
*block_info,
sendlog.info('%s blocks missing' % len(missing))
num_of_blocks = len(missing)
missing = self._upload_missing_blocks(
- missing,
- hmap,
- f,
- upload_gen)
+ missing, hmap, f, upload_gen)
if missing:
if num_of_blocks == len(missing):
retries -= 1
tries -= 1
old_failures = len(missing)
if missing:
- raise ClientError(
- '%s blocks failed to upload' % len(missing),
- details=['%s' % thread.exception for thread in missing])
+ raise ClientError('%s blocks failed to upload' % len(missing))
except KeyboardInterrupt:
sendlog.info('- - - wait for threads to finish')
for thread in activethreads():
thread.join()
raise
+ self._cb_next()
r = self.object_put(
obj,
def _dump_blocks_sync(
self, obj, remote_hashes, blocksize, total_size, dst, crange,
**args):
+ if not total_size:
+ return
for blockid, blockhash in enumerate(remote_hashes):
if blockhash:
start = blocksize * blockid
def _hash_from_file(self, fp, start, size, blockhash):
fp.seek(start)
- block = fp.read(size)
+ block = readall(fp, size)
h = newhashlib(blockhash)
h.update(block.strip('\x00'))
return hexlify(h.digest())
self._cb_next(len(blockids) - len(unsaved))
if unsaved:
key = unsaved[0]
- if key:
- self._watch_thread_limit(flying.values())
- self._thread2file(
- flying, blockid_dict, local_file, offset,
- **restargs)
- end = total_size - 1 if (
- key + blocksize > total_size) else key + blocksize - 1
- data_range = _range_up(key, end, total_size, filerange)
- if not data_range:
- self._cb_next()
- continue
- restargs[
- 'async_headers'] = {'Range': 'bytes=%s' % data_range}
- flying[key] = self._get_block_async(obj, **restargs)
- blockid_dict[key] = unsaved
+ self._watch_thread_limit(flying.values())
+ self._thread2file(
+ flying, blockid_dict, local_file, offset,
+ **restargs)
+ end = total_size - 1 if (
+ key + blocksize > total_size) else key + blocksize - 1
+ if end < key:
+ self._cb_next()
+ continue
+ data_range = _range_up(key, end, total_size, filerange)
+ if not data_range:
+ self._cb_next()
+ continue
+ restargs[
+ 'async_headers'] = {'Range': 'bytes=%s' % data_range}
+ flying[key] = self._get_block_async(obj, **restargs)
+ blockid_dict[key] = unsaved
for thread in flying.values():
thread.join()
finally:
self.container = cnt_back_up
- def get_container_info(self, until=None):
+ def get_container_info(self, container=None, until=None):
"""
:param until: (str) formated date
:raises ClientError: 404 Container not found
"""
+ bck_cont = self.container
try:
+ self.container = container or bck_cont
+ self._assert_container()
r = self.container_head(until=until)
except ClientError as err:
err.details.append('for container %s' % self.container)
raise err
+ finally:
+ self.container = bck_cont
return r.headers
def get_container_meta(self, until=None):
:returns: (dict)
"""
return filter_in(
- self.get_container_info(until=until),
- 'X-Container-Meta')
+ self.get_container_info(until=until), 'X-Container-Meta')
def get_container_object_meta(self, until=None):
"""
:returns: (dict)
"""
return filter_in(
- self.get_container_info(until=until),
- 'X-Container-Object-Meta')
+ self.get_container_info(until=until), 'X-Container-Object-Meta')
def set_container_meta(self, metapairs):
"""
finally:
from time import sleep
sleep(2 * len(activethreads()))
+ self._cb_next()
return headers.values()
def truncate_object(self, obj, upto_bytes):
:returns: (dict) response headers
"""
+ ctype = self.get_object_info(obj)['content-type']
r = self.object_post(
obj,
update=True,
content_range='bytes 0-%s/*' % upto_bytes,
- content_type='application/octet-stream',
+ content_type=ctype,
object_bytes=upto_bytes,
source_object=path4url(self.container, obj))
return r.headers
- def overwrite_object(self, obj, start, end, source_file, upload_cb=None):
+ def overwrite_object(
+ self, obj, start, end, source_file,
+ source_version=None, upload_cb=None):
"""Overwrite a part of an object from local source file
+ ATTENTION: content_type must always be application/octet-stream
:param obj: (str) remote object path
:param upload_db: progress.bar for uploading
"""
- r = self.get_object_info(obj)
- rf_size = int(r['content-length'])
- if rf_size < int(start):
- raise ClientError(
- 'Range start exceeds file size',
- status=416)
- elif rf_size < int(end):
- raise ClientError(
- 'Range end exceeds file size',
- status=416)
self._assert_container()
+ r = self.get_object_info(obj, version=source_version)
+ rf_size = int(r['content-length'])
+ start, end = int(start), int(end)
+ assert rf_size >= start, 'Range start %s exceeds file size %s' % (
+ start, rf_size)
meta = self.get_container_info()
blocksize = int(meta['x-container-block-size'])
filesize = fstat(source_file.fileno()).st_size
- datasize = int(end) - int(start) + 1
+ datasize = end - start + 1
nblocks = 1 + (datasize - 1) // blocksize
offset = 0
if upload_cb:
content_range='bytes %s-%s/*' % (
start + offset,
start + offset + len(block) - 1),
+ source_version=source_version,
data=block)
headers.append(dict(r.headers))
offset += len(block)
-
- self._cb_next
+ self._cb_next()
+ self._cb_next()
return headers
def copy_object(