Opened 3 years ago

Last modified 22 months ago

#20225 needs_information defect

tor "wedged" and become unable to connect to a hidden service

Reported by: Alan Owned by:
Priority: Medium Milestone: Tor: unspecified
Component: Core Tor/Tor Version: Tor:
Severity: Normal Keywords: tor-hs tor-client needs-diagnosis stall bootstrapping needs-insight
Cc: Actual Points:
Parent ID: Points:
Reviewer: Sponsor:


When using the latest "stable" version, tor "wedged" and become unable to connect to a hidden service. For each connection attempt, the log output printed "Rend stream is 120 seconds late. Giving up on address 'xxxxxxxxx.onion'". There was no other info in the log. Closing a restarting tor fixed the problem. I have never seen this behavior in a prior version. I decided to downgrade to v0.2.8.6 because v0.2.8.7 looked rushed and I'm not convinced it was thoroughly tested. IMO, tor should be able to solve for itself any connection problem that can be manually fixed by restarting tor.

Child Tickets

Change History (7)

comment:1 Changed 3 years ago by dgoulet

Keywords: tor-hs added
Milestone: Tor: 0.2.???
Severity: CriticalNormal
Status: newneeds_information

Thanks for this report. In order to help solve this we'll need a bit more information on this issue.

If you could provide us with Log info level of your service, it would be really appreciate it. Simply add this to your hidden service torrc: Log info file /full/path/to/log/file

If you can do the same for your tor client, great!

The main reason I suspect you failed here is because for some reasons, once you've established connection to the hidden service, the rendezvous circuit either stopped sending stuff or the hidden service application behind it failed to respond to you and the connection ultimately died because no response after 120 seconds. Logs will tell us more.

comment:2 Changed 3 years ago by Alan

I have seen not it again. If I can't reproduce it reliably, I can't give you any more info. I realize these are the most difficult bugs to fix--bugs that pop up very intermittently. They can also be the most frustrating for the user.

Regarding your theory, that doesn't explain (for me any way), why every connect attempt failed for over 5 minutes, until I finally stopped tor and restarted it, and then the problem went away.

comment:3 Changed 3 years ago by Alan

P.S., the problem was not on the hidden service side, as far as I could see. The errors were reported on the client side, and that is where I restarted tor and the problem went away. The hidden service side remained running the whole time. This is the output from the server side. I'm not sure if the two error messages that appears here are related or not:

Sep 22 18:05:07.000 [notice] Tor has successfully opened a circuit. Looks like client functionality is working.
Sep 22 18:05:07.000 [notice] Bootstrapped 100%: Done
Sep 23 00:05:05.000 [notice] Heartbeat: Tor's uptime is 5:59 hours, with 24 circuits open. I've sent 15.66 MB and received 4.35 MB.
Sep 23 04:21:26.000 [warn] connection_edge_process_relay_cell (at origin) failed.
Sep 23 06:05:05.000 [notice] Heartbeat: Tor's uptime is 11:59 hours, with 35 circuits open. I've sent 32.94 MB and received 11.04 MB.
Sep 23 09:09:56.000 [warn] connection_edge_process_relay_cell (at origin) failed.
Sep 23 12:05:05.000 [notice] Heartbeat: Tor's uptime is 17:59 hours, with 30 circuits open. I've sent 49.95 MB and received 16.48 MB.

comment:4 Changed 2 years ago by teor

Milestone: Tor: 0.2.???Tor: 0.3.???

Milestone renamed

comment:5 Changed 2 years ago by nickm

Keywords: tor-03-unspecified-201612 added
Milestone: Tor: 0.3.???Tor: unspecified

Finally admitting that 0.3.??? was a euphemism for Tor: unspecified all along.

comment:6 Changed 2 years ago by nickm

Keywords: tor-03-unspecified-201612 removed

Remove an old triaging keyword.

comment:7 Changed 22 months ago by nickm

Keywords: tor-client needs-diagnosis stall bootstrapping needs-insight added
Note: See TracTickets for help on using tickets.