summaryrefslogtreecommitdiff
path: root/crates/tor-hsservice
Commit message (Collapse)AuthorAgeFilesLines
...
* | tor-hsservice: Add an hss state mgmt APIGabriela Moldovan2023-12-142-0/+47
|/
* Merge branch 'todos-hss' into 'main'Nick Mathewson2023-12-133-25/+29
|\ | | | | | | | | Clear away some misc todos in tor-hsservice. See merge request tpo/core/arti!1819
| * hss: Explain why wait_for_netdir_to_list does not need more complexity.Nick Mathewson2023-12-131-6/+11
| |
| * hss::svc::netdir: Explain why it is okay to suppress errors.Nick Mathewson2023-12-121-1/+3
| |
| * hss::config: Downgrade or remove some TODO HSS comments.Nick Mathewson2023-12-121-16/+15
| |
| * hss: remove TODO about making DangerouslyNonAnonymous conditional.Nick Mathewson2023-12-121-2/+0
| |
* | Merge branch 'log_hsid_unconditionally' into 'main'Nick Mathewson2023-12-131-14/+20
|\ \ | | | | | | | | | | | | hsservice: Log hsid whether we just generated it or not. See merge request tpo/core/arti!1830
| * | tor-hsservice: Retrieve the public hsid for logging purposes.gabi-2502023-12-131-3/+3
| | |
| * | hsservice: Log hsid whether we just generated it or not.Nick Mathewson2023-12-131-14/+20
| |/ | | | | | | | | | | With this change you can find your hsid in your logs (if safe logging is off) even if you forgot to notice it the first time around.
* | Merge branch 'publisher-do-not-retry-fatal' into 'main'gabi-2502023-12-131-16/+1
|\ \ | | | | | | | | | | | | tor-hsservice: If the error is fatal, do not retry the desc upload. See merge request tpo/core/arti!1821
| * | tor-hsservice: If the error is fatal, do not retry the desc upload.Gabriela Moldovan2023-12-131-16/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | The errors returned by `upload_all` are now all fatal, so we there is no point in retrying `upload_all` on failure. Note this will exacerbate #1155, as it will cause the seemingly transient time skew issues to become fatal (the corresponding error type is `Bug`, so in principle they ought to be fatal)
* | | tor-hsservice: Add more explanation of replay_log lockingIan Jackson2023-12-131-0/+5
| | |
* | | tor-hsservice: Add more explanation of replay_log lockingIan Jackson2023-12-131-0/+6
| | |
* | | tor-hsservice: Cargo.toml: sortIan Jackson2023-12-131-1/+1
| | |
* | | tor-hsservice: replay: Note a bugIan Jackson2023-12-131-0/+4
| | |
* | | tor-hsservice: Make IPT logs be persistent (fmt)Ian Jackson2023-12-131-1/+3
| | |
* | | tor-hsservice: Make IPT logs be persistentIan Jackson2023-12-133-7/+88
| | |
* | | tor-hsservice ipt_mgr: Rename STORAGE_RETRY from KEYSTORE_RETRYIan Jackson2023-12-131-2/+2
| | | | | | | | | | | | We're going to reuse this for other kinds of storage error.
* | | tor-hsservice: Plumb state dir and its mistrust into ipt_mgr (fmt)Ian Jackson2023-12-131-5/+2
| | |
* | | tor-hsservice: Plumb state dir and its mistrust into ipt_mgrIan Jackson2023-12-133-3/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is needed for the replay logs. It's a shame that CheckedDir is (i) a bit unergonomic (ii) has an extra bool in it, or we could pass one of those instead of these two arguments. Since HS's might be created after startup, TorClient must have these fields.
* | | tor-hsservice tests: replay: Pass nick to ↵Ian Jackson2023-12-131-1/+2
| | | | | | | | | | | | create_storage_handles_from_state_mgr (fmt)
* | | tor-hsservice tests: replay: Pass nick to create_storage_handles_from_state_mgrIan Jackson2023-12-132-3/+6
| | | | | | | | | | | | | | | This will allow us to more faithfully model the actual Arti state directory layout.
* | | tor-hsservice tests: replay: Provide for a filesystem lockfileIan Jackson2023-12-132-2/+23
| | | | | | | | | | | | | | | This is to prevent current use of the same directory of replay logs by different instances.
* | | tor-hsservice tests: replay: Break out create_loggedIan Jackson2023-12-131-13/+25
| | |
* | | tor-hsservice: Move ReplayLog construction to ipt_mgrIan Jackson2023-12-132-9/+19
| | |
* | | tor-hsservice: Note some TODOs relating to IPT teardownIan Jackson2023-12-131-0/+9
| | |
* | | tor-hsservice: ReplayLog: Fix file magicIan Jackson2023-12-131-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This ought to have the hash algorithm name in it or we'll have trouble if we want to change the hash algorithm in the future. (Strictly, we could just choose a different magic but the string was rather short.) Add a newline, which is often convenient in file headers. And "onion" to mean "onion swervice" is improper.
* | | tor-hsservice: ReplayLog: Slight tidying upIan Jackson2023-12-131-3/+5
|/ / | | | | | | Make `#[cfg(target_family = "unix")]` appear only once.
* | tor-hsservice: Add TODO about removing the shutdown handling from the publisher.Gabriela Moldovan2023-12-131-0/+5
| |
* | tor-hsservice: Add TODO about changing a return type in ipt_set.Gabriela Moldovan2023-12-131-0/+5
| |
* | tor-hsservice: Add TODO about possibly making UploadStatus a type alias.Gabriela Moldovan2023-12-131-0/+2
| |
* | tor-hsservice: Add TODO about possibly retrying failed uploads.Gabriela Moldovan2023-12-131-0/+6
| |
* | tor-hsservice: Remove unused ReactorError type.Gabriela Moldovan2023-12-131-84/+2
| | | | | | | | | | | | | | The publisher now returns `FatalError`s, so we don't need `ReactorError` anymore. Addresses a TODO HSS in publish/reactor.rs Part of #1129
* | tor-hsservice: Replace ReactorError with FatalError (fmt).Gabriela Moldovan2023-12-132-9/+6
| |
* | tor-hsservice: Replace ReactorError with FatalError.Gabriela Moldovan2023-12-132-30/+29
| |
* | tor-hsservice: Add a NetdirProviderShutdown FatalError variant.Gabriela Moldovan2023-12-131-1/+7
| | | | | | | | This is another type of fatal error.
* | tor-hsservice: Add a FatalError::from_spawn.Gabriela Moldovan2023-12-131-0/+12
| | | | | | | | We're about to use this (in the publisher reactor).
* | tor-hsservice: Replace ReactorError::ShuttingDown with ShutdownStatus.Gabriela Moldovan2023-12-131-13/+33
| |
* | tor-hsservice: Make descriptor publisher wait for shutdown signal.Gabriela Moldovan2023-12-133-3/+29
| | | | | | | | | | | | | | | | The publisher logs a nice `info!` message when it receives the shutdown signal. The publisher can infer that the service is shutting down from the errors received on its various receiver channels (i.e. from the errors that suggest the sender was dropped), but listening for the shutdown signal is nicer.
* | tor-hsservice: Move ShutdownStatus to svc.Gabriela Moldovan2023-12-132-16/+18
| | | | | | | | | | | | | | This will be used by the descriptor publisher soon (we want to abolish its `ReactorError` altogether, and to do that, we need to get rid of `ReactorError::ShuttingDown`. `ShuttingDown::Terminate` happens to be a suitable replacement).
* | tor-hsservice: Remove unused ReactorError variant.Gabriela Moldovan2023-12-131-5/+0
| | | | | | | | | | This moves us one step closer to removing ReactorError in favour of FatalError (see the TODO HSS above ReactorError for more details).
* | tor-hsservice: Refactor upload_descriptor_with_retries to not return an error.Gabriela Moldovan2023-12-131-67/+60
| | | | | | | | | | | | | | | | | | | | | | | | Previously, this would return `ReactorError::PublishFailure` if the upload failed. However, that error wasn't used for anything other than logging. Instead of returning the error, we now log it inside `upload_descriptor_with_retries` and return an `UploadStatus` describing the upload outcome. This will enable us to abolish `ReactorError::PublishFailure` (and eventually replace `ReactorError` with `FatalError`).
* | tor-hsservice: Abolish ReactorError::HsDescBuild.Gabriela Moldovan2023-12-132-7/+3
|/ | | | | The failure to build a descriptor out of seemingly valid parts is an internal (irrecoverable) error.
* tor-hsservice: Reformat long match statement.Gabriela Moldovan2023-12-111-1/+3
|
* tor-hsservice: Reformat a long trace! statement.Gabriela Moldovan2023-12-111-1/+3
|
* tor-hsservice: Rephrase and reformat a comment.Gabriela Moldovan2023-12-111-2/+2
| | | | This line is too long.
* tor-hsservice: Remove duplicate trace! log.Gabriela Moldovan2023-12-111-5/+0
| | | | | We already log the outcome of the HsDir upload in the function that calls this code.
* tor-hsservice: Use debug! instead of trace! to log upload outcome.Gabriela Moldovan2023-12-111-1/+1
| | | | This also wraps the line.
* tor-hsservice: Remove some TODOs that have been addressed.Gabriela Moldovan2023-12-111-11/+0
|
* tor-hsservice: Publish the IptPublishSet we called note_publication_attempt on.Gabriela Moldovan2023-12-111-50/+50
| | | | | | | | | | | | | | | | | | | | | | Previously, it was possible for the `IptPublishSet` used to generate the descriptor and the `IptPublishSet` `note_publication_attempt` to differ. Now, the publisher generates the descriptor using the same `IptPublishSet` it calls `note_publication_attempt` on. Note that as a consequence, the publisher generates a new descriptor just before _each_ HsDir upload. This means each HsDir could, in theory, receive a different descriptor (not just in terms of revision-counters, but also with a different set of IPTs). It may seem like this could lead to some HsDirs being left with an outdated descriptor, but that's not the case: after the upload completes, the publisher will be notified by the ipt_watcher of the IPT change event (if there was one to begin with), which will trigger another upload job. Previously, the publisher would only generate a single descriptor for each time period (all HsDirs in a given time period would receive the same descriptor). Closes #1097
generated by cgit v1.2.3 (git 2.46.0) at 2026-08-28 19:02:12 +0000