[tor-bugs] #28639 [Core Tor/sbws]: After several days, most of the circuits timeout

[Tor Bug Tracker & Wiki]

#28639: After several days, most of the circuits timeout
------------------------------------------+--------------------------------
 Reporter:  juga                          |          Owner:  (none)
     Type:  defect                        |         Status:
                                          |  needs_information
 Priority:  Medium                        |      Milestone:  sbws:
                                          |  1.0.x-final
Component:  Core Tor/sbws                 |        Version:
 Severity:  Normal                        |     Resolution:
 Keywords:  sbws-1.0-must-moved-20181128  |  Actual Points:
Parent ID:  #25925                        |         Points:
 Reviewer:                                |        Sponsor:
------------------------------------------+--------------------------------

Comment (by teor):

 Replying to [comment:7 teor]:
 > Replying to [comment:6 juga]:
 > > Replying to [comment:1 teor]:
 > > > What does tor say in its logs?
 > >
 > > The main repeated part:
 > >
 > > {{{
 > > Oct 28 05:43:00.000 [notice] Heartbeat: Tor's uptime is 2 days 17:59
 hours, with 7 circuits open. I've sent 15.08 GB and received 482.17 GB.
 > > Oct 28 05:43:00.000 [notice] Average packaged cell fullness: 43.617%.
 TLS write overhead: 4%
 > > Oct 28 05:50:28.000 [notice] Have tried resolving or connecting to
 address '[scrubbed]' at 3 different places. Giving up.
 > > Oct 28 05:57:49.000 [notice] No circuits are opened. Relaxed timeout
 for circuit 40835 (a General-purpose client 2-hop circuit in state doing
 handshakes with channel state open) to 10000ms. However, it appears the
 circuit has timed out anyway. 0 guards are live. [53 similar message(s)
 suppressed in last 3600 seconds]
 > > }}}
 >
 > The key parts of this message are:
 > {{{
 > No circuits are opened. Relaxed timeout for circuit
 > }}}
 >
 > I think we can fix this bug by disabling adaptive timeouts in tor. See
 #28646.

 We can ignore this message. It is an unhelpful log message in Tor. It
 happens even when the circuit timeout will not change. See #28698.

 > {{{
 > 0 guards are live.
 > }}}
 >
 > > I can't figure out reasons with that, but maybe you do.
 >
 > sbws sets `UseEntryGuards 0`, so it should never have any guards:
 > https://github.com/torproject/sbws/blob/master/sbws/globals.py#L19

 Are you running Tor 0.2.9?

 The "entry guards are live" log message is not in master:
 https://github.com/torproject/tor/blob/master/src/core/or/circuituse.c#L630

 I think that it was removed in 0.3.0.1-alpha:
 https://github.com/torproject/tor/commit/68679504323b0a676a446b8fb34b976c9dc66b4f

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/28639#comment:9>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online