def extract_flow_data(out_file):
""" Given an (open) file, return a dictionary of as many elements as there are mptcp flows """
# Return at the beginning of the file
out_file.seek(0)
data = out_file.readlines()
connections = {}
current_connection = False
for line in data:
# Case 1: line start with MPTCP connection
if line.startswith("MPTCP connection"):
# A typical line: MPTCP connection 0 with id 2
words = line.split()
current_connection = int(words[-1])
connections[current_connection] = MPTCPConnection(
current_connection)
# Case 2: line for a subflow
elif current_connection is not False and line.startswith("\tSubflow"):
# A typical line:
# Subflow 0 with wscale : 6 0 IPv4 sport 59570 dport 443 saddr
# 37.185.171.74 daddr 194.78.99.114
words = line.split()
sub_flow_id = int(words[1])
subflow = MPTCPSubFlow(sub_flow_id)
index_wscale = words.index("wscale")
subflow.attr[co.WSCALESRC] = words[index_wscale + 2]
subflow.attr[co.WSCALEDST] = words[index_wscale + 3]
subflow.attr[co.TYPE] = words[index_wscale + 4]
index = words.index("sport")
while index + 1 < len(words):
attri = words[index]
value = words[index + 1]
# Cope with different format from tcptrace and mptcptrace
if attri == co.SADDR or attri == co.DADDR:
value = co.long_ipv6_address(value)
subflow.attr[attri] = value
index += 2
subflow.indicates_wifi_or_cell()
connections[current_connection].flows[sub_flow_id] = subflow
connections[current_connection].attr[co.C2S][co.BYTES] = {}
connections[current_connection].attr[co.S2C][co.BYTES] = {}
connections[current_connection].attr[co.C2S][co.RETRANS_DSS] = []
connections[current_connection].attr[co.S2C][co.RETRANS_DSS] = []
# Case 3: skip the line (no more current connection)
else:
current_connection = False
return connections
def extract_flow_data(out_file):
""" Given an (open) file, return a dictionary of as many elements as there are mptcp flows """
# Return at the beginning of the file
out_file.seek(0)
data = out_file.readlines()
connections = {}
current_connection = False
for line in data:
# Case 1: line start with MPTCP connection
if line.startswith("MPTCP connection"):
# A typical line: MPTCP connection 0 with id 2
words = line.split()
current_connection = int(words[-1])
connections[current_connection] = MPTCPConnection(current_connection)
# Case 2: line for a subflow
elif current_connection is not False and line.startswith("\tSubflow"):
# A typical line:
# Subflow 0 with wscale : 6 0 IPv4 sport 59570 dport 443 saddr
# 37.185.171.74 daddr 194.78.99.114
words = line.split()
sub_flow_id = int(words[1])
subflow = MPTCPSubFlow(sub_flow_id)
index_wscale = words.index("wscale")
subflow.attr[co.WSCALESRC] = words[index_wscale + 2]
subflow.attr[co.WSCALEDST] = words[index_wscale + 3]
subflow.attr[co.TYPE] = words[index_wscale + 4]
index = words.index("sport")
while index + 1 < len(words):
attri = words[index]
value = words[index + 1]
# Cope with different format from tcptrace and mptcptrace
if attri == co.SADDR or attri == co.DADDR:
value = co.long_ipv6_address(value)
subflow.attr[attri] = value
index += 2
subflow.indicates_wifi_or_cell()
connections[current_connection].flows[sub_flow_id] = subflow
connections[current_connection].attr[co.C2S][co.BYTES] = {}
connections[current_connection].attr[co.S2C][co.BYTES] = {}
connections[current_connection].attr[co.C2S][co.RETRANS_DSS] = []
connections[current_connection].attr[co.S2C][co.RETRANS_DSS] = []
# Case 3: skip the line (no more current connection)
else:
current_connection = False
return connections
def extract_tstat_data_tcp_complete(filename, connections, conn_id):
""" Subpart of extract_tstat_data dedicated to the processing of the log_tcp_complete file
Returns the connections seen and the conn_id reached
"""
log_file = open(filename)
data = log_file.readlines()
for line in data:
# Case 1: line start with #; skip it
if not line.startswith("#"):
# Case 2: extract info from the line
info = line.split()
conn_id += 1
connection = TCPConnection(conn_id)
connection.flow.attr[co.SADDR] = co.long_ipv6_address(info[0])
connection.flow.attr[co.DADDR] = co.long_ipv6_address(info[14])
connection.flow.attr[co.SPORT] = info[1]
connection.flow.attr[co.DPORT] = info[15]
connection.flow.detect_ipv4()
connection.flow.indicates_wifi_or_cell()
# Except RTT, all time (in ms in tstat) shoud be converted into seconds
connection.flow.attr[co.START] = float(info[28]) / 1000.0
connection.flow.attr[co.DURATION] = float(info[30]) / 1000.0
connection.flow.attr[co.C2S][co.PACKS] = int(info[2])
connection.flow.attr[co.S2C][co.PACKS] = int(info[16])
# Note that this count is about unique data bytes (sent in the payload)
connection.flow.attr[co.C2S][co.BYTES] = int(info[6])
connection.flow.attr[co.S2C][co.BYTES] = int(info[20])
# This is about actual data bytes (sent in the payload, including retransmissions)
connection.flow.attr[co.C2S][co.BYTES_DATA] = int(info[8])
connection.flow.attr[co.S2C][co.BYTES_DATA] = int(info[22])
connection.flow.attr[co.C2S][co.PACKS_RETRANS] = int(info[9])
connection.flow.attr[co.S2C][co.PACKS_RETRANS] = int(info[23])
connection.flow.attr[co.C2S][co.BYTES_RETRANS] = int(info[10])
connection.flow.attr[co.S2C][co.BYTES_RETRANS] = int(info[24])
connection.flow.attr[co.C2S][co.PACKS_OOO] = int(info[11])
connection.flow.attr[co.S2C][co.PACKS_OOO] = int(info[25])
connection.flow.attr[co.C2S][co.NB_SYN] = int(info[12])
connection.flow.attr[co.S2C][co.NB_SYN] = int(info[26])
connection.flow.attr[co.C2S][co.NB_FIN] = int(info[13])
connection.flow.attr[co.S2C][co.NB_FIN] = int(info[27])
connection.flow.attr[co.C2S][co.NB_RST] = int(info[3])
connection.flow.attr[co.S2C][co.NB_RST] = int(info[17])
connection.flow.attr[co.C2S][co.NB_ACK] = int(info[4])
connection.flow.attr[co.S2C][co.NB_ACK] = int(info[18])
# Except RTT, all time (in ms in tstat) shoud be converted into seconds
connection.flow.attr[co.C2S][co.TIME_FIRST_PAYLD] = float(
info[31]) / 1000.0
connection.flow.attr[co.S2C][co.TIME_FIRST_PAYLD] = float(
info[32]) / 1000.0
connection.flow.attr[co.C2S][co.TIME_LAST_PAYLD] = float(
info[33]) / 1000.0
connection.flow.attr[co.S2C][co.TIME_LAST_PAYLD] = float(
info[34]) / 1000.0
connection.flow.attr[co.C2S][co.TIME_FIRST_ACK] = float(
info[35]) / 1000.0
connection.flow.attr[co.S2C][co.TIME_FIRST_ACK] = float(
info[36]) / 1000.0
connection.flow.attr[co.C2S][co.RTT_SAMPLES] = int(info[48])
connection.flow.attr[co.S2C][co.RTT_SAMPLES] = int(info[55])
connection.flow.attr[co.C2S][co.RTT_MIN] = float(info[45])
connection.flow.attr[co.S2C][co.RTT_MIN] = float(info[52])
connection.flow.attr[co.C2S][co.RTT_MAX] = float(info[46])
connection.flow.attr[co.S2C][co.RTT_MAX] = float(info[53])
connection.flow.attr[co.C2S][co.RTT_AVG] = float(info[44])
connection.flow.attr[co.S2C][co.RTT_AVG] = float(info[51])
connection.flow.attr[co.C2S][co.RTT_STDEV] = float(info[47])
connection.flow.attr[co.S2C][co.RTT_STDEV] = float(info[54])
connection.flow.attr[co.C2S][co.TTL_MIN] = float(info[49])
connection.flow.attr[co.S2C][co.TTL_MIN] = float(info[56])
connection.flow.attr[co.C2S][co.TTL_MAX] = float(info[50])
connection.flow.attr[co.S2C][co.TTL_MAX] = float(info[57])
connection.flow.attr[co.C2S][co.SS_MIN] = int(info[71])
connection.flow.attr[co.S2C][co.SS_MIN] = int(info[94])
connection.flow.attr[co.C2S][co.SS_MAX] = int(info[70])
connection.flow.attr[co.S2C][co.SS_MAX] = int(info[93])
connection.flow.attr[co.C2S][co.CWIN_MIN] = int(info[76])
connection.flow.attr[co.S2C][co.CWIN_MIN] = int(info[99])
connection.flow.attr[co.C2S][co.CWIN_MAX] = int(info[75])
connection.flow.attr[co.S2C][co.CWIN_MAX] = int(info[98])
connection.flow.attr[co.C2S][co.NB_RTX_RTO] = int(info[78])
connection.flow.attr[co.S2C][co.NB_RTX_RTO] = int(info[101])
connection.flow.attr[co.C2S][co.NB_RTX_FR] = int(info[79])
connection.flow.attr[co.S2C][co.NB_RTX_FR] = int(info[102])
connection.flow.attr[co.C2S][co.NB_REORDERING] = int(info[80])
connection.flow.attr[co.S2C][co.NB_REORDERING] = int(info[103])
connection.flow.attr[co.C2S][co.NB_NET_DUP] = int(info[81])
connection.flow.attr[co.S2C][co.NB_NET_DUP] = int(info[104])
connection.flow.attr[co.C2S][co.NB_UNKNOWN] = int(info[82])
connection.flow.attr[co.S2C][co.NB_UNKNOWN] = int(info[105])
connection.flow.attr[co.C2S][co.NB_FLOW_CONTROL] = int(info[83])
connection.flow.attr[co.S2C][co.NB_FLOW_CONTROL] = int(info[106])
connection.flow.attr[co.C2S][co.NB_UNNECE_RTX_RTO] = int(info[84])
connection.flow.attr[co.S2C][co.NB_UNNECE_RTX_RTO] = int(info[107])
connection.flow.attr[co.C2S][co.NB_UNNECE_RTX_FR] = int(info[85])
connection.flow.attr[co.S2C][co.NB_UNNECE_RTX_FR] = int(info[108])
connection.attr[co.C2S][co.BYTES] = {}
connection.attr[co.S2C][co.BYTES] = {}
connection.flow.attr[co.C2S][co.TIMESTAMP_RETRANS] = []
connection.flow.attr[co.S2C][co.TIMESTAMP_RETRANS] = []
connection.flow.attr[co.C2S][co.TIME_FIN_ACK_TCP] = 0.0
connection.flow.attr[co.S2C][co.TIME_FIN_ACK_TCP] = 0.0
connection.flow.attr[co.C2S][co.TIME_LAST_ACK_TCP] = 0.0
connection.flow.attr[co.S2C][co.TIME_LAST_ACK_TCP] = 0.0
connection.flow.attr[co.C2S][co.TIME_LAST_PAYLD_TCP] = 0.0
connection.flow.attr[co.S2C][co.TIME_LAST_PAYLD_TCP] = 0.0
connection.flow.attr[co.C2S][
co.TIME_LAST_PAYLD_WITH_RETRANS_TCP] = 0.0
connection.flow.attr[co.S2C][
co.TIME_LAST_PAYLD_WITH_RETRANS_TCP] = 0.0
connections[conn_id] = connection
log_file.close()
return connections, conn_id
def extract_tstat_data_tcp_nocomplete(filename, connections, conn_id):
log_file = open(filename)
data = log_file.readlines()
for line in data:
# Case 1: line start with #; skip it
if not line.startswith("#"):
# Case 2: extract info from the line
info = line.split()
conn_id += 1
connection = TCPConnection(conn_id)
connection.flow.attr[co.SADDR] = co.long_ipv6_address(info[0])
connection.flow.attr[co.DADDR] = co.long_ipv6_address(info[14])
connection.flow.attr[co.SPORT] = info[1]
connection.flow.attr[co.DPORT] = info[15]
connection.flow.detect_ipv4()
connection.flow.indicates_wifi_or_cell()
# Except RTT, all time (in ms in tstat) shoud be converted into seconds
connection.flow.attr[co.START] = float(info[28]) / 1000.0
connection.flow.attr[co.DURATION] = float(info[30]) / 1000.0
connection.flow.attr[co.C2S][co.PACKS] = int(info[2])
connection.flow.attr[co.S2C][co.PACKS] = int(info[16])
# Note that this count is about unique data bytes (sent in the payload)
connection.flow.attr[co.C2S][co.BYTES] = int(info[6])
connection.flow.attr[co.S2C][co.BYTES] = int(info[20])
# This is about actual data bytes (sent in the payload, including retransmissions)
connection.flow.attr[co.C2S][co.BYTES_DATA] = int(info[8])
connection.flow.attr[co.S2C][co.BYTES_DATA] = int(info[22])
connection.flow.attr[co.C2S][co.PACKS_RETRANS] = int(info[9])
connection.flow.attr[co.S2C][co.PACKS_RETRANS] = int(info[23])
connection.flow.attr[co.C2S][co.BYTES_RETRANS] = int(info[10])
connection.flow.attr[co.S2C][co.BYTES_RETRANS] = int(info[24])
connection.flow.attr[co.C2S][co.PACKS_OOO] = int(info[11])
connection.flow.attr[co.S2C][co.PACKS_OOO] = int(info[25])
connection.flow.attr[co.C2S][co.NB_SYN] = int(info[12])
connection.flow.attr[co.S2C][co.NB_SYN] = int(info[26])
connection.flow.attr[co.C2S][co.NB_FIN] = int(info[13])
connection.flow.attr[co.S2C][co.NB_FIN] = int(info[27])
connection.flow.attr[co.C2S][co.NB_RST] = int(info[3])
connection.flow.attr[co.S2C][co.NB_RST] = int(info[17])
connection.flow.attr[co.C2S][co.NB_ACK] = int(info[4])
connection.flow.attr[co.S2C][co.NB_ACK] = int(info[18])
# Except RTT, all time (in ms in tstat) shoud be converted into seconds
connection.flow.attr[co.C2S][co.TIME_FIRST_PAYLD] = float(
info[31]) / 1000.0
connection.flow.attr[co.S2C][co.TIME_FIRST_PAYLD] = float(
info[32]) / 1000.0
connection.flow.attr[co.C2S][co.TIME_LAST_PAYLD] = float(
info[33]) / 1000.0
connection.flow.attr[co.S2C][co.TIME_LAST_PAYLD] = float(
info[34]) / 1000.0
connection.flow.attr[co.C2S][co.TIME_FIRST_ACK] = float(
info[35]) / 1000.0
connection.flow.attr[co.S2C][co.TIME_FIRST_ACK] = float(
info[36]) / 1000.0
connection.attr[co.C2S][co.BYTES] = {}
connection.attr[co.S2C][co.BYTES] = {}
connection.flow.attr[co.C2S][co.TIMESTAMP_RETRANS] = []
connection.flow.attr[co.S2C][co.TIMESTAMP_RETRANS] = []
connection.flow.attr[co.C2S][co.TIME_FIN_ACK_TCP] = 0.0
connection.flow.attr[co.S2C][co.TIME_FIN_ACK_TCP] = 0.0
connection.flow.attr[co.C2S][co.TIME_LAST_ACK_TCP] = 0.0
connection.flow.attr[co.S2C][co.TIME_LAST_ACK_TCP] = 0.0
connection.flow.attr[co.C2S][co.TIME_LAST_PAYLD_TCP] = 0.0
connection.flow.attr[co.S2C][co.TIME_LAST_PAYLD_TCP] = 0.0
connection.flow.attr[co.C2S][
co.TIME_LAST_PAYLD_WITH_RETRANS_TCP] = 0.0
connection.flow.attr[co.S2C][
co.TIME_LAST_PAYLD_WITH_RETRANS_TCP] = 0.0
connections[conn_id] = connection
log_file.close()
return connections, conn_id