def es_bulk_adder(worker_name, docs, cliargs, totalcrawltime=None):
starttime = time.time()
if not cliargs['s3']:
bot_logger.info('*** Bulk adding to ES index...')
try:
dirlist, filelist = docs
diskover.index_bulk_add(es, dirlist, diskover.config, cliargs)
diskover.index_bulk_add(es, filelist, diskover.config, cliargs)
except ValueError:
diskover.index_bulk_add(es, docs, diskover.config, cliargs)
if not cliargs['reindex'] and not cliargs['reindexrecurs'] and not cliargs[
'crawlbot']:
data = {
"worker_name": worker_name,
"dir_count": len(dirlist),
"file_count": len(filelist),
"bulk_time": round(time.time() - starttime, 6),
"crawl_time": round(totalcrawltime, 6),
"indexing_date": datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%S.%f")
}
es.index(index=cliargs['index'], doc_type='worker', body=data)
if not cliargs['s3']:
elapsed_time = round(time.time() - starttime, 6)
bot_logger.info('*** FINISHED BULK ADDING, Elapsed Time: ' +
str(elapsed_time))
def index_dupes(hashgroup, cliargs):
"""This is the ES dupe_md5 tag update function.
It updates a file's dupe_md5 field to be md5sum of file
if it's marked as a duplicate.
"""
bot_logger = diskover_worker_bot.bot_logger
# create Elasticsearch connection
es = diskover.elasticsearch_connect(diskover.config)
file_id_list = []
# bulk update data in Elasticsearch index
for f in hashgroup['files']:
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'file',
'_id': f['id'],
'doc': {'dupe_md5': hashgroup['md5sum']}
}
file_id_list.append(d)
if len(file_id_list) > 0:
if cliargs['verbose']:
bot_logger.info('Bulk updating %s files in ES index' % len(file_id_list))
diskover.index_bulk_add(es, file_id_list, 'file', diskover.config, cliargs)
def es_bulk_adder(result, cliargs, bot_logger):
worker_name = get_worker_name()
starttime = time.time()
dirlist = []
filelist = []
crawltimelist = []
totalcrawltime = 0
for item in result:
if item[0] == 'directory':
dirlist.append(item[1])
elif item[0] == 'file':
filelist.append(item[1])
elif item[0] == 'crawltime':
crawltimelist.append(item)
totalcrawltime += item[2]
bot_logger.info('*** Bulk adding to ES index...')
diskover.index_bulk_add(es, dirlist, 'directory', diskover.config, cliargs)
diskover.index_bulk_add(es, filelist, 'file', diskover.config, cliargs)
if not cliargs['reindex'] and not cliargs['reindexrecurs'] and not cliargs[
'crawlbot']:
diskover.add_crawl_stats_bulk(es, crawltimelist, worker_name,
diskover.config, cliargs)
data = {
"worker_name": worker_name,
"dir_count": len(dirlist),
"file_count": len(filelist),
"bulk_time": round(time.time() - starttime, 10),
"crawl_time": round(totalcrawltime, 10),
"indexing_date": datetime.utcnow().strftime("%Y-%m-%dT%H:%M:%S.%f")
}
es.index(index=cliargs['index'], doc_type='worker', body=data)
elapsed_time = round(time.time() - starttime, 3)
bot_logger.info('*** FINISHED BULK ADDING, Elapsed Time: ' +
str(elapsed_time))
def calc_dir_size(dirlist, cliargs):
"""This is the calculate directory size worker function.
It gets a directory list from the Queue search ES for all
files in each directory (recursive) and sums their filesizes
to create a total filesize and item count for each dir,
then pdates dir doc's filesize and items fields.
"""
doclist = []
for path in dirlist:
totalsize = 0
totalitems = 1 # 1 for itself
totalitems_files = 0
totalitems_subdirs = 0
# file doc search with aggregate for sum filesizes
# escape special characters
newpath = escape_chars(path[1])
# create wildcard string and check for / (root) path
if newpath == '\/':
newpathwildcard = '\/*'
else:
newpathwildcard = newpath + '\/*'
# check if / (root) path
if newpath == '\/':
data = {
"size": 0,
"query": {
"query_string": {
"query": "path_parent: " + newpath + "*",
"analyze_wildcard": "true"
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
}
}
}
else:
data = {
"size": 0,
"query": {
"query_string": {
'query': 'path_parent: ' + newpath + ' OR path_parent: ' + newpathwildcard,
'analyze_wildcard': 'true'
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
}
}
}
# search ES and start scroll
res = es.search(index=cliargs['index'], doc_type='file', body=data,
request_timeout=config['es_timeout'])
# total items sum
totalitems_files += res['hits']['total']
# total file size sum
totalsize += res['aggregations']['total_size']['value']
# directory doc search (subdirs)
# check if / (root) path
if newpath == '\/':
data = {
"size": 0,
"query": {
"query_string": {
"query": "path_parent: " + newpath + "*",
"analyze_wildcard": "true"
}
}
}
else:
data = {
"size": 0,
"query": {
"query_string": {
'query': 'path_parent: ' + newpath + ' OR path_parent: ' + newpathwildcard,
'analyze_wildcard': 'true'
}
}
}
# search ES and start scroll
res = es.search(index=cliargs['index'], doc_type='directory', body=data,
request_timeout=config['es_timeout'])
# total items sum
totalitems_subdirs += res['hits']['total']
# total items
totalitems += totalitems_files + totalitems_subdirs
# update filesize and items fields for directory (path) doc
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'directory',
'_id': path[0],
'doc': {'filesize': totalsize, 'items': totalitems,
'items_files': totalitems_files,
'items_subdirs': totalitems_subdirs}
}
# add total cost per gb to doc
if cliargs['costpergb']:
d = cost_per_gb(d, path[1], path[2], path[3], path[4], 'directory')
doclist.append(d)
index_bulk_add(es, doclist, config, cliargs)
def calc_hot_dirs(dirlist, cliargs):
"""This is the calculate hotdirs worker function.
It gets a directory list from the Queue, iterates over the path list
and searches index2 for the same path and calculates change percent
between the two. If path not in index2, change percent is 100%.
Updates index's directory doc's change_percent fields.
"""
doclist = []
for path in dirlist:
# doc search (matching path) in index2
# filename
f = os.path.basename(path[1])
# parent path
p = os.path.abspath(os.path.join(path[1], os.pardir))
data = {
"size": 1,
"_source": ['filesize', 'items', 'items_files', 'items_subdirs'],
"query": {
"query_string": {
"query": "filename: \"" + f + "\" AND path_parent: \"" + p + "\""
}
}
}
# search ES
res = es.search(index=cliargs['hotdirs'], doc_type='directory', body=data,
request_timeout=config['es_timeout'])
# calculate change percent
# set change percent to 100% if no matching path in index2
if len(res['hits']['hits']) == 0:
changepercent_filesize = 100.0
changepercent_items = 100.0
changepercent_items_files = 100.0
changepercent_items_subdirs = 100.0
else:
source = res['hits']['hits'][0]['_source']
# ((new - old) / old) * 100
try:
# check if path size in index2 was 0 bytes and set change percent to 100%
if path[2] > 0 and source['filesize'] == 0:
changepercent_filesize = 100.0
else:
changepercent_filesize = round(((path[2] - source['filesize'])
/ source['filesize']) * 100.0, 2)
except ZeroDivisionError:
changepercent_filesize = 0.0
try:
# check if path items in index2 was 0 and set change percent to 100%
if path[3] > 0 and source['items'] == 0:
changepercent_items = 100.0
else:
changepercent_items = round(((path[3] - source['items'])
/ source['items']) * 100.0, 2)
except ZeroDivisionError:
changepercent_items = 0.0
try:
# check if path file items in index2 was 0 and set change percent to 100%
if path[4] > 0 and source['items_files'] == 0:
changepercent_items_files = 100.0
else:
changepercent_items_files = round(((path[4] - source['items_files'])
/ source['items_files']) * 100.0, 2)
except ZeroDivisionError:
changepercent_items_files = 0.0
try:
# check if path subdir items in index2 was 0 and set change percent to 100%
if path[5] > 0 and source['items_subdirs'] == 0:
changepercent_items_subdirs = 100.0
else:
changepercent_items_subdirs = round(((path[5] - source['items_subdirs'])
/ source['items_subdirs']) * 100.0, 2)
except ZeroDivisionError:
changepercent_items_subdirs = 0.0
# update fields in index
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'directory',
'_id': path[0],
'doc': {'change_percent_filesize': changepercent_filesize,
'change_percent_items': changepercent_items,
'change_percent_items_files': changepercent_items_files,
'change_percent_items_subdirs': changepercent_items_subdirs}
}
doclist.append(d)
index_bulk_add(es, doclist, config, cliargs)
def tag_copier(path, cliargs):
"""This is the tag copier worker function.
It gets a path from the Queue and searches index for the
same path and copies any existing tags (from index2)
Updates index's doc's tag and tag_custom fields.
"""
bot_logger = bot_log_setup(cliargs)
jobstart = time.time()
dir_id_list = []
file_id_list = []
# doc search (matching path) in index for existing tags from index2
# filename
f = os.path.basename(path[0])
# parent path
p = os.path.abspath(os.path.join(path[0], os.pardir))
data = {
"size": 1,
"_source": ['tag', 'tag_custom'],
"query": {
"query_string": {
"query":
"filename: \"" + f + "\" AND path_parent: \"" + p + "\""
}
}
}
# check if file or directory
if path[3] == 'directory':
# search ES
res = es.search(index=cliargs['index'],
doc_type='directory',
body=data,
request_timeout=diskover.config['es_timeout'])
else:
res = es.search(index=cliargs['index'],
doc_type='file',
body=data,
request_timeout=diskover.config['es_timeout'])
# mark task done if no matching path in index and continue
if len(res['hits']['hits']) == 0:
bot_logger.info('*** No matching path found in index')
return True
# existing tag in index2
docid = res['hits']['hits'][0]['_id']
# update tag and tag_custom fields in index
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': path[3],
'_id': docid,
'doc': {
'tag': path[1],
'tag_custom': path[2]
}
}
if path[3] is 'directory':
dir_id_list.append(d)
else:
file_id_list.append(d)
diskover.index_bulk_add(es, dir_id_list, 'directory', diskover.config,
cliargs)
diskover.index_bulk_add(es, file_id_list, 'file', diskover.config, cliargs)
elapsed_time = round(time.time() - jobstart, 3)
bot_logger.info('*** FINISHED JOB, Elapsed Time: ' + str(elapsed_time))
def calc_dir_size(dirlist, cliargs):
"""This is the calculate directory size worker function.
It gets a directory list from the Queue and searches ES for all files
in each directory (recursive) and sums their filesizes
to create a total filesize and item count for each dir.
Updates dir doc's filesize and items fields.
"""
jobstart = time.time()
bot_logger.info('*** Calculating directory sizes...')
doclist = []
for path in dirlist:
totalsize = 0
totalitems = 1 # 1 for itself
totalitems_files = 0
totalitems_subdirs = 0
# file doc search with aggregate for sum filesizes
# escape special characters
newpath = diskover.escape_chars(path[1])
# create wildcard string and check for / (root) path
if newpath == '\/':
newpathwildcard = '\/*'
else:
newpathwildcard = newpath + '\/*'
# check if / (root) path
if newpath == '\/':
data = {
"size": 0,
"query": {
"query_string": {
"query": "path_parent: " + newpath + "*",
"analyze_wildcard": "true"
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
}
}
}
else:
data = {
"size": 0,
"query": {
"query_string": {
'query': 'path_parent: ' + newpath + ' OR path_parent: ' + newpathwildcard,
'analyze_wildcard': 'true'
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
}
}
}
# search ES and start scroll
res = es.search(index=cliargs['index'], doc_type='file', body=data,
request_timeout=diskover.config['es_timeout'])
# total items sum
totalitems_files += res['hits']['total']
# total file size sum
totalsize += res['aggregations']['total_size']['value']
# directory doc search (subdirs)
# search ES and start scroll
res = es.search(index=cliargs['index'], doc_type='directory', body=data,
request_timeout=diskover.config['es_timeout'])
# total items sum
totalitems_subdirs += res['hits']['total']
# total items
totalitems += totalitems_files + totalitems_subdirs
# update filesize and items fields for directory (path) doc
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'directory',
'_id': path[0],
'doc': {'filesize': totalsize, 'items': totalitems,
'items_files': totalitems_files,
'items_subdirs': totalitems_subdirs}
}
doclist.append(d)
diskover.index_bulk_add(es, doclist, diskover.config, cliargs)
elapsed_time = round(time.time() - jobstart, 3)
bot_logger.info('*** FINISHED CALC DIR, Elapsed Time: ' + str(elapsed_time))
def es_bulk_add(worker_name, dirlist, filelist, cliargs, totalcrawltime=None):
if cliargs['chunkfiles']:
updated_dirlist = []
# check for existing directory docs in index and update crawl time only (dirchunk)
for d in dirlist:
try:
path = d[
'chunkpath'] # this key determins if its part of a chunked dir
crawltime = d['crawl_time']
f = os.path.basename(path)
# parent path
p = os.path.abspath(os.path.join(path, os.pardir))
data = {
"size": 1,
"_source": ['crawl_time'],
"query": {
"query_string": {
"query":
"filename: \"" + f + "\" AND path_parent: \"" + p +
"\""
}
}
}
es.indices.refresh(index=cliargs['index'])
res = es.search(index=cliargs['index'],
doc_type='directory',
body=data,
request_timeout=config['es_timeout'])
if len(res['hits']['hits']) == 0:
continue
docid = res['hits']['hits'][0]['_id']
current_crawltime = res['hits']['hits'][0]['_source'][
'crawl_time']
udpated_crawltime = current_crawltime + crawltime
# update crawltime in index
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'directory',
'_id': docid,
'doc': {
'crawl_time': udpated_crawltime
}
}
except KeyError:
pass # not part of a chunked dir
updated_dirlist.append(d)
dirlist = updated_dirlist
starttime = time.time()
docs = dirlist + filelist
index_bulk_add(es, docs, config, cliargs)
if not cliargs['noworkerdocs']:
data = {
"worker_name": worker_name,
"dir_count": len(dirlist),
"file_count": len(filelist),
"bulk_time": round(time.time() - starttime, 6),
"crawl_time": round(totalcrawltime, 6),
"indexing_date": datetime.utcnow().isoformat()
}
es.index(index=cliargs['index'], doc_type='worker', body=data)
def calc_dir_size(dirlist, cliargs):
"""This is the calculate directory size worker function.
It gets a directory list from the Queue and searches ES for all
subdirs in each directory (recursive) and sums their filesize and
items fields to create a total filesize and item count for each directory doc.
Updates directory doc's filesize and items fields.
"""
# check if other bots are idle and throw them some jobs (dir paths)
if len(dirlist) >= cliargs['batchsize']:
workers_idle = 0
workers = Worker.all(connection=redis_conn)
num_workers = len(workers)
for w in workers:
if w._state == "idle":
workers_idle += 1
if workers_idle > num_workers // 2:
workers_idle = True
break
q_len = len(q_calc)
if q_len == 0 and workers_idle == True:
# take half the paths randomly
shuffle(dirlist)
n = len(dirlist) // 2
tossdirs = dirlist[:n]
dirlist = dirlist[n:]
q_crawl.enqueue(calc_dir_size, args=(
tossdirs,
cliargs,
))
doclist = []
for path in dirlist:
totalitems = 1 # 1 for itself
# file doc search with aggregate for sum filesizes
# escape special characters
newpath = escape_chars(path[1])
parentpath = escape_chars(
os.path.abspath(os.path.join(path[1], os.pardir)))
pathbasename = escape_chars(os.path.basename(path[1]))
# create wildcard string and check for / (root) path
if newpath == '\/':
newpathwildcard = '\/*'
else:
newpathwildcard = newpath + '\/*'
# check if / (root) path
if newpath == '\/':
data = {
"size": 0,
"query": {
"query_string": {
"query": "path_parent: " + newpath + "*",
"analyze_wildcard": "true"
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
}
}
}
else:
data = {
"size": 0,
"query": {
"query_string": {
'query':
'(path_parent: ' + parentpath + ' AND filename: ' +
pathbasename + ') OR path_parent: ' + newpath +
' OR path_parent: ' + newpathwildcard,
'analyze_wildcard':
'true'
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
},
"total_files": {
"sum": {
"field": "items_files"
}
},
"total_subdirs": {
"sum": {
"field": "items_subdirs"
}
}
}
}
# search ES and start scroll for all directory doc search (subdirs)
res = es.search(index=cliargs['index'],
doc_type='directory',
body=data,
request_timeout=config['es_timeout'])
# total file size sum
totalsize = res['aggregations']['total_size']['value']
# total items sum for all subdirs count
totalitems_subdirs = res['aggregations']['total_subdirs']['value']
# total items sum for all files count
totalitems_files = res['aggregations']['total_files']['value']
totalitems += totalitems_subdirs + totalitems_files
# update filesize and items fields for directory (path) doc
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'directory',
'_id': path[0],
'doc': {
'filesize': totalsize,
'items': totalitems,
'items_files': totalitems_files,
'items_subdirs': totalitems_subdirs
}
}
doclist.append(d)
index_bulk_add(es, doclist, config, cliargs)
def calc_dir_size(dirlist, cliargs):
"""This is the calculate directory size worker function.
It gets a directory list from the Queue and searches ES for all
subdirs in each directory (recursive) and sums their filesize and
items fields to create a total filesize and item count for each directory doc.
Updates directory doc's filesize and items fields.
"""
doclist = []
for path in dirlist:
totalitems = 1 # 1 for itself
# file doc search with aggregate for sum filesizes
# escape special characters
newpath = escape_chars(path[1])
parentpath = escape_chars(
os.path.abspath(os.path.join(path[1], os.pardir)))
pathbasename = escape_chars(os.path.basename(path[1]))
# create wildcard string and check for / (root) path
if newpath == '\/':
newpathwildcard = '\/*'
else:
newpathwildcard = newpath + '\/*'
# check if / (root) path
if newpath == '\/':
data = {
"size": 0,
"query": {
"query_string": {
"query": "path_parent: " + newpath + "*",
"analyze_wildcard": "true"
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
}
}
}
else:
data = {
"size": 0,
"query": {
"query_string": {
'query':
'(path_parent: ' + parentpath + ' AND filename: ' +
pathbasename + ') OR path_parent: ' + newpath +
' OR path_parent: ' + newpathwildcard,
'analyze_wildcard':
'true'
}
},
"aggs": {
"total_size": {
"sum": {
"field": "filesize"
}
},
"total_files": {
"sum": {
"field": "items_files"
}
},
"total_subdirs": {
"sum": {
"field": "items_subdirs"
}
}
}
}
# search ES and start scroll for all directory doc search (subdirs)
res = es.search(index=cliargs['index'],
doc_type='directory',
body=data,
request_timeout=config['es_timeout'])
# total file size sum
totalsize = res['aggregations']['total_size']['value']
# total items sum for all subdirs count
totalitems_subdirs = res['aggregations']['total_subdirs']['value']
# total items sum for all files count
totalitems_files = res['aggregations']['total_files']['value']
totalitems += totalitems_subdirs + totalitems_files
# update filesize and items fields for directory (path) doc
d = {
'_op_type': 'update',
'_index': cliargs['index'],
'_type': 'directory',
'_id': path[0],
'doc': {
'filesize': totalsize,
'items': totalitems,
'items_files': totalitems_files,
'items_subdirs': totalitems_subdirs
}
}
doclist.append(d)
index_bulk_add(es, doclist, config, cliargs)
