• Welcome to Hurricane Electric's IPv6 Tunnel Broker Forums.

News:

Welcome to Hurricane Electric's Tunnelbroker.net forums!

Main Menu

Usage statistics

Started by ndm, May 13, 2008, 05:58:30 AM

Previous topic - Next topic

ndm

Hi

how often are the statistics updated? They don't seem to be "live".


Regards André

samh

They are updating right now, and will be running at roughly this time each morning, they only check for ipv6 latency from your tunnel endpoint so if you dont allow icmp from your tunnel server you wont ever show up on those stats.

snarked

Is the latency list being updated for the new tunnels added?

For example, the Los Angeles tunnel is listing only 7 (for a top 20) but there are more than 50 tunnels allocated, so shouldn't the list be a full 20 users long?

Miami lists only one entry, but (as of the time of this post) has 58 tunnels.

samh

Not necessarily, Ill have a look at the automation scripts but the test only attempts to generate latency information to the client v6 address.  Lots of people dont respond to ping6 etc.. so they dont show up in the list.

Sam


Quote from: snarked on August 05, 2008, 01:07:25 PM
Is the latency list being updated for the new tunnels added?

For example, the Los Angeles tunnel is listing only 7 (for a top 20) but there are more than 50 tunnels allocated, so shouldn't the list be a full 20 users long?

Miami lists only one entry, but (as of the time of this post) has 58 tunnels.

snarked

Quote... Lots of people dont respond to ping6 etc.. so they dont show up in the list.
Aren't they supposed to, else the tunnel is considered inactive and can be temporarily closed (like when someone who only has a dynamic connection is offline)?

snarked

Not only do the latency statistics for Los Angeles appear unchanged - they actually appear static in that the TTLs reported have not changed in 2 weeks.  Also, there are now 63 tunnels allocated and I doubt that all of the new tunnels are inactive.  I don't think that the new tunnels are in your update routine/script.

kriteknetworks

Been wondering about this myself, my ping time to ashburn has been 32.534ms for almost two weeks now heh

snarked

Los Angeles appears fixed (or at least it updated today).  However, Miami still has a single entry for t#16 and only 5 for t#12 - but has 5+68 for number of tunnels allocated.  Maybe that location's statistics are still broken.

snarked

cf. http://www.tunnelbroker.net/usage/tunnels_by_latency.php

Since it is permitted for an account to have more than one tunnel at a given tunnel server (provided they have different IPv4 endpoints), it would be nice to know WHICH tunnel is being listed in the statistics.

To please those who might not want the fact of more than one tunnel disclosed, I also suggest an account setting to allow or suppress the tunnel # (i.e. 1-5) to be suffixed to the [user] name displayed in the list.  If implemented, default it to suppress.

Today, tunnel server #4 (New York) has one user listed twice (at positions 9 and 18).  How does that user know which tunnel is which (without this suggestion)?

Of course, this should only happen when a user has more than one tunnel (through the same tunnel server).  Otherwise, there is no confusion and thus no need to identify a tunnel number when there's only one.

kcochran

I went with the simple route.

If you're logged in, you'll see the associated tunnel id with your own tunnels, if you're on the list.  If you're not logged in, you won't see anything different from now.  In any case, you won't see anyone else's tunnel IDs.

snarked

Close enough.  Thanks.