Commit graph

1939 commits

Author SHA1 Message Date
Roger Dingledine
db34f6c9e8 handle other de-sync cases from #40396
Specifically, every time a guard moves into or out of state
GUARD_REACHABLE_MAYBE, it is an opportunity for the guard reachability
state to get out of sync with the have-minimum-dir-info state.

Fixes even more of #40396.
2021-10-28 20:57:28 -04:00
Roger Dingledine
3c8510e2c0 reassess minimum-dir-info when a bridge fails
When we try to fetch a bridge descriptor and we fail, we mark
the guard as failed, but we never scheduled a re-compute for
router_have_minimum_dir_info().

So if we had already decided we needed to wait for this new descriptor,
we would just wait forever -- even if, counterintuitively, *losing* the
bridge is just what we need to *resume* using the network, if we had it
in state GUARD_REACHABLE_MAYBE and we were stalling to learn this outcome.

See bug 40396 for more details.
2021-10-28 20:57:28 -04:00
Roger Dingledine
867c3c6f89 only log "new bridge descriptor" if really new
The bridge descriptor fetching codes ends up fetching a lot of duplicate
bridge descriptors, because this is how we learn when the descriptor
changes.

This commit only changes comments plus whether we log that one line.

It moves us back to the old behavior, before the previous commit for
30496, where we would only log that line when the bridge descriptor
we're talking about is better than the one we already had (if any).
2021-10-28 20:57:28 -04:00
Roger Dingledine
d66549c208 fetch missing bridge descriptors without delay
Without this change, if we have a working bridge, and we add a new bridge,
we will schedule the fetch attempt for that new bridge descriptor for
three hours(!) in the future.

This change is especially needed because of bug #40396, where if you have
one working bridge and one bridge whose descriptor you haven't fetched
yet, your Tor will stall until you have successfully fetched that new
descriptor -- in this case for hours.

In the old design, we would put off all further bridge descriptor fetches
once we had any working bridge descriptor. In this new design, we make the
decision per bridge based on whether we successfully got *its* descriptor.

To make this work, we need to also call learned_bridge_descriptor() every
time we get a bridge descriptor, not just when it's a novel descriptor.

Fixes bug 40396.

Also happens to fix bug 40495 (redundant descriptor fetches for every
bridge) since now we delay fetches once we succeed.

A side effect of this change is that if we have any configured bridges
that *aren't* working, we will keep trying to fetch their descriptors
on the modern directory retry schedule -- every couple of seconds for
the first half minute, then backing off after that -- which is a lot
faster than before.
2021-10-24 17:40:28 -04:00
Alexander Færøy
ae05f06597 Merge branch 'tor-gitlab/mr/452_squashed' into main 2021-10-21 12:57:37 +00:00
Nick Mathewson
ab26475cab Add a new consensus method to handle MiddleOnly specially.
When this method is in place, then any relay which is assigned
MiddleOnly has Exit, V2Dir, Guard, and HSDir cleared
(and has BadExit set if appropriate).
2021-10-21 12:57:20 +00:00
Nick Mathewson
fc542167cb Implement a MiddleOnly flag for vote generation.
This proposal implements part of Prop335; it's based on a patch
from Neel Chauhan.

When configured to do so, authorities will assign a MiddleOnly flag
to certain relays.  Any relay which an authority gives this flag
will not get Exit, V2Dir, Guard, or HSDir, and might get BadExit if
the authority votes for that one.
2021-10-21 12:57:20 +00:00
Alexander Færøy
d320f4d2a2 Merge remote-tracking branch 'tor-gitlab/mr/442' into main 2021-10-21 12:50:28 +00:00
David Goulet
7c2c749d89 relay: Comment out a unused variable for now
We keep it around until libevent is fixed, it should be used again. In
the meantime, avoid the compiler to complain of this unused variable.

https://gitlab.torproject.org/dgoulet/tor/-/jobs/43358#L1522

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-20 15:17:08 -04:00
David Goulet
903fb3dd62 relay: Avoid duplicate MetricsPort DNS error
We don't output per-type DNS errors anymore so avoid looping over the
DNS query type and output each errors for them. Before this commit, it
created 3x the same message because we had A, AAAA and PTR type records.

Fix on previous commit e7abab8782

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-20 13:49:44 -04:00
David Goulet
e7abab8782 relay: For metrics, don't report DNS errors by query type
This is due to the libevent bug
https://github.com/libevent/libevent/issues/1219 that fails to return
back the DNS record type on error.

And so, the MetricsPort now only reports the errors as a global counter
and not a per record type.

Closes #40490

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-20 10:40:56 -04:00
David Goulet
7a8108ea87 relay: Overload state on DNS timeout is now X% over Y secs
With this commit, we will only report a general overload state if we've
seen more than X% of DNS timeout errors over Y seconds. Previous
behavior was to report when a single timeout occured which is really too
small of a threshold.

The value X is a consensus parameters called
"overload_dns_timeout_scale_percent" which is a scaled percentage
(factor of 1000) so we can represent decimal points for X like 0.5% for
instance. Its default is 1000 which ends up being 1%.

The value Y is a consensus parameters called
"overload_dns_timeout_period_secs" which is the time period for which
will gather DNS errors and once over, we assess if that X% has been
reached ultimately triggering a general overload signal.

Closes #40491

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-20 10:00:03 -04:00
David Goulet
2a705e81a3 Merge branch 'maint-0.4.6' 2021-10-19 10:35:40 -04:00
David Goulet
f748a497c7 Merge branch 'ticket40476_045_01' into ticket40476_046_01 2021-10-19 09:48:13 -04:00
David Goulet
db297a177e hs: Improve warning for bad service version
Now that we don't have version 2, it gives us:

  [warn] HiddenServiceVersion must be between 3 and 3, not 2.

This commit changes it to:

  [warn] HiddenServiceVersion must be 3, not 2.

Part of #40476

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-19 09:36:14 -04:00
David Goulet
044eadae65 hs-v2: Disable version 2 HSPOST and HSFETCH command
Part of #40476

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-19 09:20:53 -04:00
David Goulet
a7d28da272 hs-v2: Disable version 2 directory
Relay do not accept both stores and lookups of version 2 descriptor.
This effectively disable version 2 HSDir supports for relays.

Part of #40476

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-19 09:16:13 -04:00
David Goulet
e284b9f779 hs-v2: Disable version 2 introduction point
Upon receiving a v2 introduction request, the relay will close the
circuit and send back a tor protocol error.

Part of #40476

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-19 09:13:08 -04:00
David Goulet
471149b651 hs-v2: Disable version 2 service
The minimum service version is raised from 2 to 3 which effectively
disable loading or creating an onion service v2.

As for ADD_ONION, for version 2, a 551 error is returned:

  "551 Failed to add Onion Service"

Part of #40476

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-19 09:11:12 -04:00
Nick Mathewson
34f3529861 Merge branch 'hs_fuzzing' 2021-10-16 10:49:41 -04:00
Nick Mathewson
37c0542d0d Merge remote-tracking branch 'asn-private/hsfuz' into hs_fuzzing 2021-10-15 11:20:06 -04:00
Alexander Færøy
8c18e9a949 Merge remote-tracking branch 'tor-gitlab/mr/459' into main 2021-10-14 19:19:32 +00:00
Alexander Færøy
b93af906c4 Merge remote-tracking branch 'tor-gitlab/mr/456' into main 2021-10-14 19:18:22 +00:00
Nick Mathewson
15ede0435f Lower maximum value for guard-extreme-restriction-percent to 100.
Values greater than 100 would have had the same effect as 100, so
this doesn't actually change Tor's behavior; it just makes the
intent clearer.  Fixes #40486; see also torspec#66.
2021-10-14 12:39:05 -04:00
Nick Mathewson
d10ceb7165 Downgrade "Rejecting RENDEZVOUS1 cell with unrecognized cookie"
This is the loudest of our LOG_PROTOCOL_WARN messages, it can occur
naturally, and there doesn't seem to be a great response to it.

Partial fix for 40400; bugfix on 0.1.1.13-alpha.
2021-10-14 12:21:30 -04:00
David Goulet
e0a6a0d085 hs: Fix memory leak if service failed to configure
Closes #40484

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-14 10:15:16 -04:00
David Goulet
0f7e0d5f2f dirauth: Reject EOL relays
Series 0.4.2.x, 0.4.3.x and 0.4.4.x will all be rejected at the
authority level at this commit.

Futhermore, the 0.4.5.x alphas and rc will also be rejected.

Closes #40480

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-10-08 11:09:33 -04:00
David Goulet
adcb094cb6 Merge branch 'tor-gitlab/mr/392' into maint-0.4.5 2021-10-06 15:45:13 -04:00
David Goulet
065ebd10c2 Merge branch 'tor-gitlab/mr/393' into maint-0.4.5 2021-10-06 15:41:12 -04:00
David Goulet
474c85a98d Merge branch 'maint-0.4.5' into maint-0.4.6 2021-10-06 15:35:43 -04:00
Mike Perry
6dae9903b1 Turn orconn watermarks into consensus parameters.
Tuning these may reduce memory usage and latency.
2021-10-04 10:45:46 -04:00
Mike Perry
0422eb26a7 Prop#324: Hook up flow control 2021-10-04 10:45:46 -04:00
Neel Chauhan
b7992d4f79 bwauth: Add AuthDirDontVoteOnDirAuthBandwidth option to avoid giving weights to dirauths 2021-09-27 08:58:00 -07:00
Neel Chauhan
8ead53330c Add spaces between the "and" when logging the "Your server has not managed to confirm reachability for its" on dual-stack relays 2021-08-26 13:40:53 -07:00
Nick Mathewson
984e3a9c6c Merge remote-tracking branch 'tor-gitlab/mr/420' 2021-08-18 08:43:31 -04:00
David Goulet
cac612af42 dir: Do not flag non-running failing HSDir
When a directory request fails, we flag the relay as non Running so we
don't use it anymore.

This can be problematic with onion services because there are cases
where a tor instance could have a lot of services, ephemeral ones, and
keeps failing to upload descriptors, let say due to a bad network, and
thus flag a lot of nodes as non Running which then in turn can not be
used for circuit building.

This commit makes it that we never flag nodes as non Running on a onion
service directory request (upload or fetch) failure as to keep the
hashring intact and not affect other parts of tor.

Fortunately, the onion service hashring is _not_ selected by looking at
the Running flag but since we do a 3-hop circuit to the HSDir, other
services on the same instance can influence each other by removing nodes
from the consensus for path selection.

This was made apparent with a small network that ran out of nodes to
used due to rapid succession of onion services uploading and failing.
See #40434 for details.

Fixes #40434

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-08-17 13:27:14 -04:00
David Goulet
70d8fb3eab relay: Reduce streaming compression ratio from HIGH to LOW
Fixes #40301

Signed-off-by: David Goulet <dgoulet@torproject.org>
2021-08-11 12:47:20 +00:00
George Kadianakis
5949a70b58 Add stricter limits to the number of L2 nodes 2021-07-12 12:27:25 +03:00
George Kadianakis
7f701794e7 Add a switch to toggle the feature on/off 2021-07-12 12:22:58 +03:00
George Kadianakis
9b464cdc36 Add layer2_guard_free() 2021-07-09 17:06:26 +03:00
George Kadianakis
a77727cdca Don't double-pick L2 nodes 2021-07-09 17:06:22 +03:00
George Kadianakis
1f87269cf4 Code improvements 2021-07-06 13:22:59 +03:00
George Kadianakis
e23947716e Use L2 vanguards during path selection
Co-authored-by: Mike Perry <mikeperry-git@torproject.org>
2021-07-01 18:15:55 +03:00
George Kadianakis
314a6b42c5 Introduce vanguards-lite subsystem and some of its entry points
Co-authored-by: Mike Perry <mikeperry-git@torproject.org>
2021-07-01 18:15:55 +03:00
Nick Mathewson
e71db3a4be Merge remote-tracking branch 'tor-gitlab/mr/405' 2021-07-01 09:56:35 -04:00
David Goulet
e9edcea0ca Merge branch 'tor-gitlab/mr/275' 2021-06-29 10:55:46 -04:00
Steven Engler
f944e46399 If TestingTorNetwork, skip perm check on the hs directory 2021-06-28 13:51:06 -04:00
George Kadianakis
2b97c1dd34 Merge remote-tracking branch 'tor-gitlab/mr/385' 2021-06-28 14:11:46 +03:00
David Goulet
270398fa31 Merge branch 'maint-0.4.6' 2021-06-18 14:29:15 -04:00
Alexander Færøy
2c00ad36cd Fix compilation on systems with older compilers.
This patch fixes a build error with GCC 7.x which doesn't seem to accept
const int's as constants in macro initialization.

See: tpo/core/tor#40410
2021-06-18 18:14:07 +00:00