> a tor op :
> Hi
> When a tor admin updates a tor node, what is the reasoning for
> punishing the status by removing flags like the guard flag?
It's not a punishment, it's an advisory notice to clients not to choose
the relay as a guard (or HSDir, or relay for Stable circuits).
The Tor network prioritises client circuit reliability and latency, not
relay bandwidth usage. This makes the user experience better.
If a relay's version becomes too old, it will be removed from the
network. But that's not a punishment either, it's done to protect
clients, so that users can have a secure and reliable tor network.
> The node may have been up for months on end without issues and goes
> down for a few minutes during install and restart and comes up with a
> newer version, hence it is clearly updated. And the guard flag goes
> away. Doesn't seem really appropriate.
A period of downtime is often correlated with another subsequent period
of downtime. But if a relay stays stable for a few days or weeks, then
it's much less likely it will go down again.
> Unless it's to indicate caution
> due to new version perhaps not being stable.
People run unstable versions on the tor network to help us find bugs.
We really appreciate this, because not all bugs show up in testing.
> But then that's what you
> do pre-prod testing in-house for.
We do integration testing, and pre-prod testing in-house.
But we have nothing that's the scale or age of the current tor network.
For more information, please feel free to search the list archives.
(This question is asked reasonably often.)
T
--
Tim Wilson-Brown (teor)
teor2345 at gmail dot com
PGP C855 6CED 5D90 A0C5 29F6 4D43 450C BA7F 968F 094B
ricochet:ekmygaiu4rzgsk6n
xmpp: teor at torproject dot org
signature.asc
Description: Message signed with OpenPGP
___
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
