Add script to upload packages to web server (#13545)

- Why I did it
The flow of SONiC reproducible build assumes that web packages (retrieved by curl and wget) are located in a single web file.
server. Currently there is no way to easily download the packages from the various file servers where they are currently deployed.
In this change we offer a utility to download all relevant packages and upload them to a specific file server.

- How I did it
We implemented python script that parse the various version files (generated by SONIC reproducible build compilation) and identify all relevant packages.
Later the script downloads the file and then upload them to the destination server pointed by the user.

- How to verify it
Script is running and was verified manually

Which release branch to backport (provide reason below if selected)
Feature will be added to master only

Signed-off-by: oreiss <oren.reiss@gmail.com>
This commit is contained in:
Oren Reiss 2023-02-23 08:18:42 +02:00 committed by GitHub
parent d15f5201e9
commit 729eb07d5f
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -0,0 +1,295 @@
#!/usr/bin/env python3
import sys
import os
import time
import argparse
from http import HTTPStatus
try:
import requests
except ImportError:
print("requests module is not installed. script will fail to execute")
#debug print level
PRINT_LEVEL_ERROR = "err"
PRINT_LEVEL_WARN = "warn"
PRINT_LEVEL_INFO = "info"
PRINT_LEVEL_VERBOSE = "verbose"
PRINT_LEVEL_LUT = {PRINT_LEVEL_ERROR : 1,
PRINT_LEVEL_WARN : 2,
PRINT_LEVEL_INFO : 3,
PRINT_LEVEL_VERBOSE : 4 }
#return code
RET_CODE_SUCCESS = 0
RET_CODE_CANNOT_CREATE_FILE = -1
RET_CODE_CANNOT_OPEN_FILE = -2
RET_CODE_HTTP_SERVER_ERROR = -3
RET_CODE_CANNOT_WRITE_FILE = -4
#constants
RESOURCES_FILE_NAME = 'versions-web'
EXCLUDE_DIRECTORES = ['fsroot', 'target']
HASH_SEPARATOR = '-'
DEFAULT_INVALID_INPUT = 'none'
# global variables
g_current_print_level = PRINT_LEVEL_INFO
#Script debug features (disabled by default)
g_delete_resources_in_cache = True
# global Classes
class Resource:
def __init__(self, line, file):
self.file = file
temp=line.split("==")
assert(2==len(temp))
self.url=temp[0].strip()
self.hash=temp[1].strip()
temp=self.url.split("/")
assert(len(temp)>0)
self.name=temp[len(temp)-1]
#handle special scenarios
if 0 != self.name.count('?') == True:
temp = self.name.split("?")
self.name = temp[0]
def get_unique_name(self):
return self.name + HASH_SEPARATOR + self.hash
def get_url(self):
return self.url
def __str__(self):
ret_val = "Resource name: " + self.name + "\n"
ret_val += "File: " + self.file + "\n"
ret_val += "Hash: " + self.hash + "\n"
ret_val += "Full URL: " + self.url
return ret_val
# Helper functions
def print_msg(print_level, msg, print_in_place=False):
if PRINT_LEVEL_LUT[g_current_print_level] >= PRINT_LEVEL_LUT[print_level]:
if True == print_in_place:
print(msg, end='\r')
else:
print(msg)
def create_dir_if_not_exist(dir):
if not os.path.exists(dir):
try:
os.makedirs(dir)
except:
print_msg(PRINT_LEVEL_WARN, "Cannot create directory " + dir)
def delete_file_if_exist(file):
if os.path.exists(file):
try:
os.remove(file)
except:
print_msg(PRINT_LEVEL_WARN, "Cannot delete " + file)
# Logic functions
def generate_output_file(resources, dest_url_valid, dest_url, output_file_name):
try:
with open(output_file_name, 'w') as f:
for unique_name in resources.keys():
resource = resources[unique_name]
if True == dest_url_valid:
line = dest_url
else:
line = resource.get_url()
if line[-1] != '/':
line += '/'
line += resource.name + "==" + resource.hash
f.write(line + '\n')
except:
print_msg(PRINT_LEVEL_WARN, output_file_name + " cannot be created")
return RET_CODE_CANNOT_CREATE_FILE
return RET_CODE_SUCCESS
def upload_resource_to_server(resource_path, resource_name, user, key, server_url):
url_full_path = server_url + "/" + resource_name
try:
f = open(resource_path, 'rb')
except:
err_print("Cannot open " + resource_path)
return RET_CODE_CANNOT_OPEN_FILE
headers = {'Content-type': 'application', 'Slug': resource_name}
response = requests.put(url_full_path, data=f,
headers=headers, auth=(user, key))
f.close()
if response.status_code != HTTPStatus.CREATED.value:
err_print(f"HTTP request returned status code {response.status_code}, expected {HTTPStatus.CREATED.value}")
return RET_CODE_HTTP_SERVER_ERROR
# JSON response empty only when status code is 204
reported_md5 = response.json().get('checksums', {}).get('md5')
file_md5 = resource_name.split(HASH_SEPARATOR)[-1]
# Check if server reports checksum, if so compare reported sum and the one
# specified in filename
if reported_md5 != None and reported_md5 != file_md5:
print_msg(PRINT_LEVEL_WARN, f"Server reported file's chsum {reported_md5}, expected {file_md5}")
return RET_CODE_SUCCESS
def download_external_resouce(resource, cache_path):
resource_path_in_cache = cache_path + os.sep + resource.get_unique_name()
r = requests.get(resource.get_url(), allow_redirects=True)
try:
f = open(resource_path_in_cache, 'wb')
f.write(r.content)
f.close()
except:
print_msg(PRINT_LEVEL_ERROR, "Cannot write " + resource_path_in_cache + " to cache")
resource_path_in_cache = "" #report error
return resource_path_in_cache
def get_resources_list(resource_files_list):
resource_list = list()
for file_name in resource_files_list:
try:
with open(file_name, 'r') as f:
for line in f:
resource_list.append(Resource(line, file_name))
except:
print_msg(PRINT_LEVEL_WARN, file_name + " cannot be opened")
return resource_list
def filter_out_dir(subdir):
ret_val = False
for exclude in EXCLUDE_DIRECTORES:
if exclude in subdir.split(os.sep):
ret_val = True
break
return ret_val
def get_resource_files_list(serach_path):
resource_files_list = list()
for subdir, dirs, files in os.walk(serach_path):
for file in files:
if False == filter_out_dir(subdir) and RESOURCES_FILE_NAME == file:
file_full_path = os.path.join(subdir, file)
print_msg(PRINT_LEVEL_VERBOSE, "Found resource file :" + file_full_path)
resource_files_list.append(file_full_path)
return resource_files_list
def parse_args():
parser = argparse.ArgumentParser(description='Various pre-steps for build compilation')
parser.add_argument('-s', '--source', default=".",
help='Search path for ' + RESOURCES_FILE_NAME + ' files')
parser.add_argument('-c', '--cache', default="." + os.sep + "tmp",
help='Path to cache for storing content before uploading to server')
parser.add_argument('-p', '--print', default=PRINT_LEVEL_INFO,
choices=[PRINT_LEVEL_ERROR, PRINT_LEVEL_WARN, PRINT_LEVEL_INFO, PRINT_LEVEL_VERBOSE],
help='Print level verbosity')
parser.add_argument('-o', '--output', default=DEFAULT_INVALID_INPUT,
help='Output file name to hold the list of packages')
parser.add_argument('-u', '--user', default=DEFAULT_INVALID_INPUT,
help='User for server authentication')
parser.add_argument('-k', '--key', default=DEFAULT_INVALID_INPUT,
help='API key server authentication')
parser.add_argument('-d', '--dest', default=DEFAULT_INVALID_INPUT,
help='URL for destination web file server')
return parser.parse_args()
def main():
global g_current_print_level
ret_val = RET_CODE_SUCCESS
resource_counter = 0.0
resource_dict = dict()
args = parse_args()
g_current_print_level = args.print
resource_files_list = get_resource_files_list(args.source)
resource_list = get_resources_list(resource_files_list)
#remove duplications
for resource in resource_list:
unique_name = resource.get_unique_name()
if not unique_name in resource_dict.keys():
resource_dict[unique_name] = resource
print_msg(PRINT_LEVEL_INFO, "Found " + str(len(resource_files_list)) + " version files and " + str(len(resource_dict.keys())) + " unique resources")
if args.dest != DEFAULT_INVALID_INPUT:
upload_files_to_server = True
print_msg(PRINT_LEVEL_INFO, "Upload files to URL - " + args.dest)
else:
upload_files_to_server = False
print_msg(PRINT_LEVEL_INFO, "Skipping files upload to server")
#create cache directory if not exist
create_dir_if_not_exist(args.cache)
#download content to cache and then upload to web server
for unique_name in resource_dict.keys():
resource = resource_dict[unique_name]
print_msg(PRINT_LEVEL_VERBOSE, resource)
resource_counter += 1.0
#download content to cache
file_in_cache = download_external_resouce(resource, args.cache)
if "" == file_in_cache:
return RET_CODE_CANNOT_WRITE_FILE
if True == upload_files_to_server:
#upload content to web server
ret_val = upload_resource_to_server(file_in_cache, unique_name, args.user, args.key, args.dest)
if ret_val != RET_CODE_SUCCESS:
return ret_val
if True == g_delete_resources_in_cache:
delete_file_if_exist(file_in_cache)
print_msg(PRINT_LEVEL_INFO, "Downloading Data. Progress " + str(int(100.0*resource_counter/len(resource_dict.keys()))) + "%", True) #print progress bar
# generate version output file as needed
if args.output != DEFAULT_INVALID_INPUT:
ret_val = generate_output_file(resource_dict, upload_files_to_server, args.dest, args.output)
print_msg(PRINT_LEVEL_INFO, "Generate output file " + args.output)
return ret_val
# Entry function
if __name__ == '__main__':
ret_val = main()
sys.exit(ret_val)