Still had 1 address policies cached at shutdown

changed milestone to %Tor: 0.3.2.x-final

added component::core tor/tor milestone::Tor: 0.3.2.x-final priority::low resolution::fixed severity::minor status::closed type::defect version::tor 0.3.0.10 labels

Trac:
Priority: Medium to Low
Severity: Normal to Minor
Milestone: N/A to Tor: 0.3.2.x-final

May 07 22:16:25.499 [warn] Still had 12 address policies cached at shutdown.
May 07 22:16:25.499 [warn]   1 [27]: reject 107.181.174.84:*
May 07 22:16:25.499 [warn]   2 [1239]: reject *4:*
May 07 22:16:25.499 [warn]   3 [45]: reject 10.0.0.0/8:*
May 07 22:16:25.499 [warn]   4 [27]: accept *4:443
May 07 22:16:25.499 [warn]   5 [18]: accept *4:*
May 07 22:16:25.499 [warn]   6 [45]: reject 172.16.0.0/12:*
May 07 22:16:25.499 [warn]   7 [45]: reject 127.0.0.0/8:*
May 07 22:16:25.499 [warn]   8 [45]: reject 169.254.0.0/16:*
May 07 22:16:25.499 [warn]   9 [27]: accept *4:80
May 07 22:16:25.499 [warn]   10 [45]: reject 192.168.0.0/16:*

Confirmed that bug appears consistently on Tor 0.3.0.6 as well.

Trac:
Version: N/A to Tor: 0.3.0.6

I just saw this on a dirauth running 0.3.0.6 when killing it:

May 16 09:32:26.000 [notice] Catching signal TERM, exiting cleanly.
May 16 09:32:27.000 [warn] Still had 10 address policies cached at shutdown.
May 16 09:32:27.000 [warn]   1 [3]: reject 107.181.174.84:*
May 16 09:32:27.000 [warn]   2 [3]: reject 169.254.0.0/16:*
May 16 09:32:27.000 [warn]   3 [3]: reject 10.0.0.0/8:*
May 16 09:32:27.000 [warn]   4 [3]: reject 127.0.0.0/8:*
May 16 09:32:27.000 [warn]   5 [3]: accept *4:443
May 16 09:32:27.000 [warn]   6 [3]: reject 192.168.0.0/16:*
May 16 09:32:27.000 [warn]   7 [3]: reject 172.16.0.0/12:*
May 16 09:32:27.000 [warn]   8 [3]: accept *4:80
May 16 09:32:27.000 [warn]   9 [3]: reject 0.0.0.0/8:*
May 16 09:32:27.000 [warn]   10 [170]: reject *4:*

Here's our hint:

$ grep 107.181.174.84 cached-*
cached-consensus:r NeelTorExitUSWest 1GF1SHwzLMYC99GkTXunM0jGxJ4 oJ8kgsI8e732ZTLPiAklOcWJxyA 2017-05-21 02:38:29 107.181.174.84 443 80
cached-microdesc-consensus:r NeelTorExitUSWest 1GF1SHwzLMYC99GkTXunM0jGxJ4 2017-05-21 02:38:29 107.181.174.84 443 80

It's in the consensus but moria1 doesn't have a descriptor for it!

Here's something interesting in my info-level log:

May 21 01:10:40.108 [info] dirserv_router_get_status(): Descriptor from router $D46175487C332CC602F7D1A44D7BA73348C6C49E~NeelTorExitUSWest at 107.181.174.84 has an Ed25519 key, but the <rsa,ed25519> keys don't match what they were before.

In fact, I have an info-level log line like that once a minute in my log, forever.

Replying to arma:

It's in the consensus but moria1 doesn't have a descriptor for it! [...] In fact, I have an info-level log line like that once a minute in my log, forever.

I've opened that one as #22349 (moved). It might be related to this ticket but it's not the same.

Wonder if #22370 (moved) is related.

Closing since #22370 (moved) looks like the root cause.

Trac:
Status: new to closed
Resolution: N/A to duplicate

I have seen this on a modified 0.3.0.10 with the #22370 (moved) patch (but also a bunch of extra privcount patches, none of which deal with address policies).

Is this still happening on an unmodified master?

Trac:
Status: closed to reopened
Resolution: duplicate to N/A
Version: Tor: 0.3.0.6 to Tor: 0.3.0.10
Milestone: Tor: 0.3.2.x-final to Tor: 0.3.3.x-final

Trac:
000a.zip

zip of authority data dir with this bug

I don't think I've been seeing this one lately. Did we fix it in some other bugfix, and this is now a duplicate of whatever-ticket-that-was?

I haven't seen it on 0.3.2.9, so let's call it fixed in 0.3.2 or 0.3.1

Trac:
Status: reopened to closed
Milestone: Tor: 0.3.3.x-final to Tor: 0.3.2.x-final
Resolution: N/A to fixed

closed

mentioned in issue #22370 (moved)

moved to tpo/core/tor#21926 (closed)

Still had 1 address policies cached at shutdown

Child items ...

Activity