最近有需求就是冷儲存資料進行異地災備，同時為了更多的節省本地的儲存成本，維護成本，人力資源等等，選擇使用相對更為優惠的百度物件儲存來進行備份資料，BOS產品介紹：BOS介紹，為了快速的，批量的上傳檔案，利用BOS Python SDK開發了一套分散式多工上傳解決方案，本文主要來介紹一下BOS Python SDK的使用方法，為BOS免費做了廣告，怎麼感謝我！

一、建立虛擬環境

# yum install python-virtualenv
# mkvirtualenv bos
# workon bos

列出虛擬環境列表：

workon/lsvirtualenv

新建虛擬環境：

mkvirtualenv [虛擬環境名稱]

啟動/切換虛擬環境：

workon [虛擬環境名稱]

刪除虛擬環境：
rmvirtualenv [虛擬環境名稱]

離開虛擬環境：

deactive

二、安裝BOS SDK

2.1 下載bos sdk安裝包

wget http://sdk.bce.baidu.com/console-sdk/bce-python-sdk-0.8.8.zip

python setup.py install

三、編寫配置檔案

bos_sample_conf.py：

import logging
import os
import sys
from baidubce.bce_client_configuration import BceClientConfiguration
from baidubce.auth.bce_credentials import BceCredentials

PROXY_HOST = 'localhost:8080'                                                                                                                                                            
bos_host = "bj.bcebos.com"
access_key_id = "a6748c1334a44c2d8af60fcdf098b30d"
secret_access_key = "3d7621d35b0c426ea2c0dfdbfca45151"

logger = logging.getLogger('baidubce.services.bos.bosclient')
fh = logging.FileHandler("sample.log")
fh.setLevel(logging.DEBUG)

formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
fh.setFormatter(formatter)
logger.setLevel(logging.DEBUG)
logger.addHandler(fh)

config = BceClientConfiguration(credentials=BceCredentials(access_key_id, secret_access_key), endpoint = bos_host)
 

四、檔案上傳

import os,sys,hashlib
from baidubce import exception
from baidubce.services import bos
from baidubce.services.bos import canned_acl
from baidubce.services.bos.bos_client import BosClient
import base64
import bos_sample_conf

##init a bos client
bos_client = BosClient(bos_sample_conf.config)

##init a bucket
bucket_name = 'wahaha'
if not bos_client.does_bucket_exist(bucket_name):
    bos_client.create_bucket(bucket_name)
    print "init bucket:%s success" % bucket_name
##upload object from string
object_key = 'Happy Spring Festival'
str = 'this is the test string'
bos_client.put_object_from_string(bucket_name,object_key,str)

##put object from file
file_name = "/root/baidu_object_storage/test/file_to_be_upload"
response = bos_client.put_object_from_file(bucket_name, object_key + ' plus',file_name)
print "response.metadata.etag = " + response.metadata.etag
##get object meta data
response = bos_client.get_object_meta_data(bucket_name, object_key+' plus')
print "response object meta data:" 
print response

##list objects in bucket
response = bos_client.list_objects(bucket_name)
for object in response.contents:
    print 'object.key = ' + object.key

##get bucket list
response = bos_client.list_buckets()
for bucket in response.buckets:
    print "bucket.name = " + bucket.name
##get object
print bos_client.get_object_as_string(bucket_name,object_key)

#get unfinished multipart upload task
print "get unfinished multipart upload task:"
for item in bos_client.list_all_multipart_uploads(bucket_name):
    print 'item.upload_id = ' + item.upload_id

#abort unfinished multipart upload task
print "abort unfinished multipart upload task"
for item in bos_client.list_all_multipart_uploads(bucket_name):
    bos_client.abort_multipart_upload(bucket_name, item.key.encode("utf-8"), upload_id = item.upload_id)

response = bos_client.list_multipart_uploads(bucket_name)
for item in response.bukcet:
    print item.name
 

五、大檔案上傳

import os
import sys
import hashlib

sys.path.append("../bos")
import bos_sample_conf
from baidubce import exception
from baidubce.services import bos
from baidubce.services.bos import canned_acl
from baidubce.services.bos.bos_client import BosClient


default_path = os.path.dirname(os.path.realpath(__file__))

#init a bos client
bos_client = BosClient(bos_sample_conf.config)

#init a bucket
bucket_name = 'wahaha'
if not bos_client.does_bucket_exist(bucket_name):
    bos_client.create_bucket(bucket_name)

#init object key
object_key = 'this is object_key of big file'

#upload multipart object
upload_id = bos_client.initiate_multipart_upload(bucket_name,object_key).upload_id
print 'upload_id = ' + upload_id
file_name = default_path + os.path.sep + 'big_file'
if os.path.isfile(file_name):
    print "file_name = %s" % file_name
else:
    exit(-1)
#set the beginning of multipart
left_size = os.path.getsize(file_name)
#set the offset 
offset = 0
part_number = 1
part_list = []
e_tag_str = ""

while left_size > 0:
    #set each part 50MB
    print "size left: %dMB" %  (left_size/1014/1024)
    part_size = 50*1024*1024
    if left_size < part_size:
        part_size = left_size
    
    response = bos_client.upload_part_from_file(
            bucket_name,object_key,upload_id,part_number,part_size,file_name,offset)

    left_size -= part_size
    offset += part_size
    part_list.append({
        "partNumber":part_number,
        "eTag":response.metadata.etag
        })
    part_number += 1
    e_tag_str += response.metadata.etag
    print part_number, " ", response.metadata.etag


print "\n"
response = bos_client.complete_multipart_upload(bucket_name,object_key,upload_id,part_list)
print response

m = hashlib.md5()
m.update(e_tag_str)
e_tag_str_to_md5 = "-" + m.hexdigest()

if e_tag_str_to_md5 == response.etag:
    print "e_tag match great!!!"
else:
    print "etag does not match, e_tag_str_to_md5 = %s" % e_tag_str_to_md5

print "\n"
print response.bucket
print response.key
print response.etag
print response.location

注：

1.對於小檔案上傳返回的response值的etag值就是小檔案的md5值。

2.對於大檔案上傳，完成分塊上傳的etag是每一個分塊上傳返回的etag值相加之後再計算md5值得到的值之前再加個"-"（好抽象）。

Author：憶之獨秀

Email：[email protected]