Loss of Guard and HS Dir flags
I currently run 3 relays, across different servers and today I noticed that one has now lost it's Guard and HS Dir flags. What's surprising is that this particular relay has the highest Bandwidth and Consensus Weight of all 3 and has not been restarted for over a month. The stats for all 3 can be found with: OhNoAnotherRelay I know that just running the relay is the important part and caring about what flags it has is a side issue, but I'm just interested in seeing why this has happened. Cheers.
On 10/13/2021 11:29 PM, Eddie wrote:
I currently run 3 relays, across different servers and today I noticed that one has now lost it's Guard and HS Dir flags. What's surprising is that this particular relay has the highest Bandwidth and Consensus Weight of all 3 and has not been restarted for over a month.
The stats for all 3 can be found with: OhNoAnotherRelay
I know that just running the relay is the important part and caring about what flags it has is a side issue, but I'm just interested in seeing why this has happened.
Cheers.
Now the relay has lost the Long-Lived Circuits flag despite showing an up-time of 34 days. There's something strange going on here. Cheers.
All: Are the DirectoryAuthority & FallbackDir directives only evaluated at startup of a Tor instance? I recently ran into an issue where my Tor Relay Farm went down, due to the manually configured DirectoryAuthority going down and the FallbackDir didn't seem to back it up. I know that the DirectoryAuthority & FallbackDir directives were intended to be used in a private Tor network, but as I'm running a Tor Relay Farm where all Tor Relay Nodes use a cloned .tordb–I need all the Tor Relay Nodes to talk to the same DirectoryAuthority & FallbackDir. Other than the DirectoryAuthority going down and a Medium-Term key update, loadbalancing to a Tor Relay Farm has been working very well for a month or so. Although, the fundamental Tor Relay client-based push reporting vs DirectoryAuthority server-based pull reporting doesn't lend itself well to accurately reporting Tor Metrics for a Tor Relay Farm. The DirectoryAuthority observers the Tor Farm based on the individual Tor Relay Node that last reported, so Tor Metrics are only shown for a single Tor Relay Node; while, in reality the Tor Relay Farm is doing x4 - x5 the reported load. Knowing this, you might reconsider the way metrics are reported in the future to accurately report loadbalanced Tor Relays. As always, thank you for all the work you do. Respectfully, Gary— This Message Originated by the Sun. iBigBlue 63W Solar Array (~12 Hour Charge) + 2 x Charmast 26800mAh Power Banks = iPhone XS Max 512GB (~2 Weeks Charged)
David Goulet or Georg Koppen: Can either of you confirm my question regarding whether the DirectoryAuthority & FallbackDir directives only evaluated at startup of a Tor instance? I need to verify the best way to configure cloned, loadbalanced Tor Relay Instances to interface with DirectoryAuthorities. Presently, I have 5 Asuswrt RT-AC66U_B1 (2 Core CPU + 256MB) WiFi Routers (1 as Nginx Loadbalancer + 4 AiMesh Nodes) schizophrenicly configured as cloned Tor Relay Nodes pushing ~175GB (~10,000 circuits) of Tor traffic, daily, which is working very well for such a small footprint. With this final piece sorted, I'll be able to contribute a tutorial for loadbalancing Tor Relays, which can be extrapolated to larger, high-availability architectural implementations that appear as a single Tor Node (opposed to having several Tor Nodes that comprise a family). Your feedback is greatly appreciated. Respectfully, Gary— This Message Originated by the Sun. iBigBlue 63W Solar Array (~12 Hour Charge) + 2 x Charmast 26800mAh Power Banks = iPhone XS Max 512GB (~2 Weeks Charged) On Monday, October 18, 2021, 2:10:12 AM MDT, Gary C. New via tor-relays <tor-relays@lists.torproject.org> wrote: All: Are the DirectoryAuthority & FallbackDir directives only evaluated at startup of a Tor instance? I recently ran into an issue where my Tor Relay Farm went down, due to the manually configured DirectoryAuthority going down and the FallbackDir didn't seem to back it up. I know that the DirectoryAuthority & FallbackDir directives were intended to be used in a private Tor network, but as I'm running a Tor Relay Farm where all Tor Relay Nodes use a cloned .tordb–I need all the Tor Relay Nodes to talk to the same DirectoryAuthority & FallbackDir. Other than the DirectoryAuthority going down and a Medium-Term key update, loadbalancing to a Tor Relay Farm has been working very well for a month or so. Although, the fundamental Tor Relay client-based push reporting vs DirectoryAuthority server-based pull reporting doesn't lend itself well to accurately reporting Tor Metrics for a Tor Relay Farm. The DirectoryAuthority observers the Tor Farm based on the individual Tor Relay Node that last reported, so Tor Metrics are only shown for a single Tor Relay Node; while, in reality the Tor Relay Farm is doing x4 - x5 the reported load. Knowing this, you might reconsider the way metrics are reported in the future to accurately report loadbalanced Tor Relays. As always, thank you for all the work you do. Respectfully, Gary— This Message Originated by the Sun. iBigBlue 63W Solar Array (~12 Hour Charge) + 2 x Charmast 26800mAh Power Banks = iPhone XS Max 512GB (~2 Weeks Charged) _______________________________________________ tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
Eddie:
On 10/13/2021 11:29 PM, Eddie wrote:
I currently run 3 relays, across different servers and today I noticed that one has now lost it's Guard and HS Dir flags. What's surprising is that this particular relay has the highest Bandwidth and Consensus Weight of all 3 and has not been restarted for over a month.
The stats for all 3 can be found with: OhNoAnotherRelay
I know that just running the relay is the important part and caring about what flags it has is a side issue, but I'm just interested in seeing why this has happened.
Cheers.
Now the relay has lost the Long-Lived Circuits flag despite showing an up-time of 34 days.
You mean the Stable flag? Seems to be back now at least. I've filed a ticket[1], though, to look a bit closer at what is going on but am not sure how fast we get to that (help is welcomed!).
There's something strange going on here.
Maybe. :) Georg [1] https://gitlab.torproject.org/tpo/network-health/team/-/issues/128
On 10/18/2021 1:03 AM, Georg Koppen wrote:
Eddie:
On 10/13/2021 11:29 PM, Eddie wrote:
I currently run 3 relays, across different servers and today I noticed that one has now lost it's Guard and HS Dir flags. What's surprising is that this particular relay has the highest Bandwidth and Consensus Weight of all 3 and has not been restarted for over a month.
The stats for all 3 can be found with: OhNoAnotherRelay
I know that just running the relay is the important part and caring about what flags it has is a side issue, but I'm just interested in seeing why this has happened.
Cheers. Now the relay has lost the Long-Lived Circuits flag despite showing an up-time of 34 days. You mean the Stable flag? Seems to be back now at least. I've filed a ticket[1], though, to look a bit closer at what is going on but am not sure how fast we get to that (help is welcomed!). And the Stable flag has gone again. There's something strange going on here. Maybe. :)
Georg
[1] https://gitlab.torproject.org/tpo/network-health/team/-/issues/128
Necro alert !! Again, the same server has exhibited the same thing. It has lost the Guard, HS Dir, and Long-Lived (Stable) flags, despite being up for almost 5 months with no reduction in Bandwidth. Cheers. On 10/18/2021 9:42 PM, Eddie wrote:
On 10/18/2021 1:03 AM, Georg Koppen wrote:
Eddie:
On 10/13/2021 11:29 PM, Eddie wrote:
I currently run 3 relays, across different servers and today I noticed that one has now lost it's Guard and HS Dir flags. What's surprising is that this particular relay has the highest Bandwidth and Consensus Weight of all 3 and has not been restarted for over a month.
The stats for all 3 can be found with: OhNoAnotherRelay
I know that just running the relay is the important part and caring about what flags it has is a side issue, but I'm just interested in seeing why this has happened.
Cheers. Now the relay has lost the Long-Lived Circuits flag despite showing an up-time of 34 days. You mean the Stable flag? Seems to be back now at least. I've filed a ticket[1], though, to look a bit closer at what is going on but am not sure how fast we get to that (help is welcomed!). And the Stable flag has gone again. There's something strange going on here. Maybe. :)
Georg
[1] https://gitlab.torproject.org/tpo/network-health/team/-/issues/128
_______________________________________________ tor-relays mailing list tor-relays@lists.torproject.org https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays
---------------------------------------------------------------------- This e-mail was checked for spam by the freeware edition of CleanMail. The freeware edition is restricted to personal and non-commercial use. You can remove this notice by purchasing a commercial license: http://antispam.byteplant.com/products/cleanmail/index.html
Eddie <stunnel@attglobal.net> wrote:
Necro alert !!
Again, the same server has exhibited the same thing.? It has lost the Guard, HS Dir, and Long-Lived (Stable) flags, despite being up for almost 5 months with no reduction in Bandwidth.
Cheers.
On 10/18/2021 9:42 PM, Eddie wrote:
On 10/18/2021 1:03 AM, Georg Koppen wrote:
Eddie:
On 10/13/2021 11:29 PM, Eddie wrote:
I currently run 3 relays, across different servers and today I noticed that one has now lost it's Guard and HS Dir flags.? What's surprising is that this particular relay has the highest Bandwidth and Consensus Weight of all 3 and has not been restarted for over a month.
The stats for all 3 can be found with:? OhNoAnotherRelay
I know that just running the relay is the important part and caring about what flags it has is a side issue, but I'm just interested in seeing why this has happened.
Cheers. Now the relay has lost the Long-Lived Circuits flag despite showing an up-time of 34 days. You mean the Stable flag? Seems to be back now at least. I've filed a ticket[1], though, to look a bit closer at what is going on but am not sure how fast we get to that (help is welcomed!). And the Stable flag has gone again. There's something strange going on here. Maybe. :)
Yes, there definitely is, as I have been pointing out repeatedly on this list over the last several years. The Authority servers *used* to deal out these flags rationally and consistently before that time. Since then they have acted capriciously w.r.t. to those flags. BTW, you omitted the Fast flag, which has also been subject to such mishandling.
Georg
[1] https://gitlab.torproject.org/tpo/network-health/team/-/issues/128
Scott Bennett, Comm. ASMELG, CFIAG ********************************************************************** * Internet: bennett at sdf.org *xor* bennett at freeshell.org * *--------------------------------------------------------------------* * "A well regulated and disciplined militia, is at all times a good * * objection to the introduction of that bane of all free governments * * -- a standing army." * * -- Gov. John Hancock, New York Journal, 28 January 1790 * **********************************************************************
participants (4)
-
Eddie -
Gary C. New -
Georg Koppen -
Scott Bennett