[or-cvs] r19751: {torflow} Change how we record the timestamp to be per-node, based on (torflow/trunk/NetworkScanners/BwAuthority)
mikeperry at seul.org
mikeperry at seul.org
Sun Jun 21 02:48:02 UTC 2009
Author: mikeperry
Date: 2009-06-20 22:48:02 -0400 (Sat, 20 Jun 2009)
New Revision: 19751
Modified:
torflow/trunk/NetworkScanners/BwAuthority/aggregate.py
Log:
Change how we record the timestamp to be per-node, based on
selection. Also change the measuring to round to nearest
1000 in addition to 3 sig figs. Also, fix a div by 1024 bug.
Modified: torflow/trunk/NetworkScanners/BwAuthority/aggregate.py
===================================================================
--- torflow/trunk/NetworkScanners/BwAuthority/aggregate.py 2009-06-21 02:43:55 UTC (rev 19750)
+++ torflow/trunk/NetworkScanners/BwAuthority/aggregate.py 2009-06-21 02:48:02 UTC (rev 19751)
@@ -15,6 +15,7 @@
nodes = {}
prev_consensus = {}
ALPHA = 0.3333 # Prev consensus values count for 1/3 of the avg
+MAX_AGE = 60*60*24*4 # Discard measurements from more than 2 days ago
def base10_round(bw_val):
# This keeps the first 3 decimal digits of the bw value only
@@ -23,7 +24,8 @@
if bw_val == 0:
plog("NOTICE", "Zero bandwidth!")
return 0
- return int(round(bw_val,-(int(math.log10(bw_val))-2)))
+ return int(max((1000,
+ round(round(bw_val,-(int(math.log10(bw_val))-2)), -3))))
def closest_to_one(ratio_list):
min_dist = 0x7fffffff
@@ -46,6 +48,7 @@
self.node_data = {}
self.idhex = None
self.nick = None
+ self.chosen_time = None
self.chosen_sbw = None
self.chosen_fbw = None
self.sbw_ratio = None
@@ -55,6 +58,7 @@
self.strm_bw = []
self.filt_bw = []
self.ns_bw = []
+ self.timestamps = []
def add_line(self, line):
if self.idhex and self.idhex != line.idhex:
@@ -65,6 +69,8 @@
or self.node_data[line.slice_file].timestamp < line.timestamp:
self.node_data[line.slice_file] = NodeData(line.timestamp)
+ # FIXME: This is kinda nutty. Can we simplify? For instance,
+ # do these really need to be lists inside the nd?
nd = self.node_data[line.slice_file]
nd.strm_bw.append(line.strm_bw)
nd.filt_bw.append(line.filt_bw)
@@ -73,11 +79,14 @@
self.strm_bw = []
self.filt_bw = []
self.ns_bw = []
+ self.timestamps = []
for nd in self.node_data.itervalues():
self.strm_bw.extend(nd.strm_bw)
self.filt_bw.extend(nd.filt_bw)
self.ns_bw.extend(nd.ns_bw)
+ for i in xrange(len(nd.ns_bw)):
+ self.timestamps.append(nd.timestamp)
def avg_strm_bw(self):
return sum(self.strm_bw)/float(len(self.strm_bw))
@@ -110,6 +119,7 @@
def main(argv):
TorUtil.read_config(argv[1]+"/scanner.1/bwauthority.cfg")
+ TorUtil.loglevel = "INFO"
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect((TorUtil.control_host,TorUtil.control_port))
@@ -156,16 +166,15 @@
slicenum = sr+"/"+fp.readline()
timestamp = float(fp.readline())
fp.close()
+ if time.time() - timestamp > MAX_AGE:
+ plog("INFO", "Skipping old file "+f)
+ continue
bw_files.append((slicenum, timestamp, sr+"/"+f))
+ # FIXME: Can we kill this?
if slicenum not in timestamps or \
timestamps[slicenum] < timestamp:
timestamps[slicenum] = timestamp
- # FIXME: Hrmm.. there may be edge cases here where we have
- # an extra slice number that is never scanned again (ie the network
- # shrinks). This will leave us with really old timestamps.
- oldest_timestamp = min(timestamps.itervalues())
-
# Need to only use most recent slice-file for each node..
for (s,t,f) in bw_files:
fp = file(f, "r")
@@ -189,28 +198,46 @@
pre_filt_avg = sum(map(lambda n: n.avg_filt_bw(), nodes.itervalues()))/ \
float(len(nodes))
+ plog("INFO", "Network pre_strm_avg: "+str(pre_strm_avg))
+ plog("INFO", "Network pre_filt_avg: "+str(pre_filt_avg))
+
for n in nodes.itervalues():
n.choose_strm_bw(pre_strm_avg)
n.choose_filt_bw(pre_filt_avg)
+ plog("DEBUG", "Node "+n.nick+" chose sbw: "+\
+ str(n.strm_bw[n.chosen_sbw])+" fbw: "+\
+ str(n.filt_bw[n.chosen_fbw]))
true_strm_avg = sum(map(lambda n: n.strm_bw[n.chosen_sbw],
nodes.itervalues()))/float(len(nodes))
true_filt_avg = sum(map(lambda n: n.filt_bw[n.chosen_fbw],
nodes.itervalues()))/float(len(nodes))
+ plog("INFO", "Network true_strm_avg: "+str(true_strm_avg))
+ plog("INFO", "Network true_filt_avg: "+str(true_filt_avg))
+
for n in nodes.itervalues():
n.fbw_ratio = n.filt_bw[n.chosen_fbw]/true_filt_avg
n.sbw_ratio = n.strm_bw[n.chosen_sbw]/true_strm_avg
if closest_to_one((n.sbw_ratio, n.fbw_ratio)) == 0:
n.ratio = n.sbw_ratio
n.new_bw = n.ns_bw[n.chosen_sbw]*n.ratio
+ n.chosen_time = n.timestamps[n.chosen_sbw]
else:
n.ratio = n.fbw_ratio
n.new_bw = n.ns_bw[n.chosen_fbw]*n.ratio
+ n.chosen_time = n.timestamps[n.chosen_fbw]
+ # XXX: Bytes or kb, we really need to decide.. and be careful when
+ # we do
if n.idhex in prev_consensus and prev_consensus[n.idhex].bandwidth != None:
prev_consensus[n.idhex].measured = True
- n.new_bw = ((prev_consensus[n.idhex].bandwidth*ALPHA + n.new_bw)/(ALPHA + 1))/1024.0
+ n.new_bw = ((prev_consensus[n.idhex].bandwidth*ALPHA + n.new_bw)/(ALPHA + 1))
+ oldest_timestamp = min(map(lambda n: n.chosen_time,
+ filter(lambda n: n.idhex in prev_consensus,
+ nodes.itervalues())))
+ plog("INFO", "Oldest measured node: "+time.ctime(oldest_timestamp))
+
for n in prev_consensus.itervalues():
if not n.measured:
if "Fast" in n.flags and "Running" in n.flags:
@@ -231,7 +258,7 @@
out = file(argv[-1], "w")
out.write(str(int(round(oldest_timestamp,0)))+"\n")
for n in n_print:
- out.write("node_id="+n.idhex+" bw="+str(base10_round(n.new_bw))+" nick="+n.nick+"\n")
+ out.write("node_id="+n.idhex+" bw="+str(base10_round(n.new_bw))+" nick="+n.nick+" measured_at="+str(n.chosen_time)+"\n")
out.close()
if __name__ == "__main__":
More information about the tor-commits
mailing list