Opened 5 years ago

Closed 5 years ago

#1706 closed defect (fixed)

Participant Bandwidth Reporting Double-Counted

Reported by: Obscuratus Owned by: zzz
Priority: minor Milestone: 0.9.23
Component: apps/console Version: 0.9.22
Keywords: Cc:
Parent Tickets: Sensitive: no


The Bandwidth reporting for Participant tunnels appears to be double-counting the bandwidth of the active peers shown in the table.

The bandwidth for the participant tunnels seems to be accumulated in the lifetime bandwidth. But it is added in a second time in (see patch below).

When a particularly high BW peer drops off the table after the grace period expires, you will see the overall bandwidth actually appear to decrease.

This has been previously (mis-)reported on the forum as "Fake participating traffic"


But, it isn't "fake" traffic, it is merely due to a peer that was being double-counted expiring after it's grace period.

NOTE: Apologies in advance if I don't reply to comments. I don't think I have permissions to comment yet.

From c57f884c37b488079ede75a426aa57a817a10496 Mon Sep 17 00:00:00 2001
From: obscuratus <obscuratus@mail.i2p>
Date: Sun, 1 Nov 2015 10:55:09 -0600
Subject: [PATCH] RFC i2p.router.web: Adjust Participant Bandwidth Reporting

The Participating Tunnels Lifetime Bandwidth Usage appears to be
double-counting the bandwidth of the active participants.

Signed-off-by: obscuratus <obscuratus@mail.i2p>
 apps/routerconsole/java/src/net/i2p/router/web/ | 1 -
 1 file changed, 1 deletion(-)

diff --git a/apps/routerconsole/java/src/net/i2p/router/web/ b/apps/routerconsole/java/src/net/i2p/router/web/
index a761a5a..f89413c 100644
--- a/apps/routerconsole/java/src/net/i2p/router/web/
+++ b/apps/routerconsole/java/src/net/i2p/router/web/
@@ -86,7 +86,6 @@ public class TunnelRenderer {
-            processed += count;
             if (++displayed > DISPLAY_LIMIT)


Change History (3)

comment:1 Changed 5 years ago by zzz

Component: unspecifiedapps/console
Milestone: undecided0.9.23
Owner: set to zzz
Status: newaccepted

Agreed with your analysis.
Permissions should be fixed, we had you in there with a lower-case 'o'.

comment:2 Changed 5 years ago by zzz

The problem is much worse. The participatingMessageCount stat is an internal per-tunnel-lifetime average used for RouterThrottle? and has little relation to lifetime total. Will split into two stats.

comment:3 Changed 5 years ago by zzz

Resolution: fixed
Status: acceptedclosed
Note: See TracTickets for help on using tickets.