Commit Graph

356 Commits

Author SHA1 Message Date
realbigsean
7a71977987 Clippy 1.49.0 updates and dht persistence test fix (#2156)
## Issue Addressed

`test_dht_persistence` failing

## Proposed Changes

Bind `NetworkService::start` to an underscore prefixed variable rather than `_`.  `_` was causing it to be dropped immediately

This was failing 5/100 times before this update, but I haven't been able to get it to fail after updating it

Co-authored-by: realbigsean <seananderson33@gmail.com>
2021-01-19 00:34:28 +00:00
realbigsean
423dea169c update smallvec (#2152)
## Issue Addressed

`cargo audit` is failing because of a potential for an overflow in the version of `smallvec` we're using

## Proposed Changes

Update to the latest version of `smallvec`, which has the fix


Co-authored-by: realbigsean <seananderson33@gmail.com>
2021-01-11 23:32:11 +00:00
Age Manning
2931b05582 Update libp2p (#2101)
This is a little bit of a tip-of-the-iceberg PR. It houses a lot of code changes in the libp2p dependency. 

This needs a bit of thorough testing before merging. 

The primary code changes are:
- General libp2p dependency update
- Gossipsub refactor to shift compression into gossipsub providing performance improvements and improved API for handling compression



Co-authored-by: Paul Hauner <paul@paulhauner.com>
2020-12-23 07:53:36 +00:00
Pawan Dhananjay
f998eff7ce Subnet discovery fixes (#2095)
## Issue Addressed

N/A

## Proposed Changes

Fixes multiple issues related to discovering of subnet peers.
1. Subnet discovery retries after yielding no results
2. Metadata updates if peer send older metadata
3. peerdb stores the peer subscriptions from gossipsub
2020-12-17 00:39:15 +00:00
Age Manning
dfb588e521 Softer penalties for missing blocks (#2075)
## Issue Addressed

Users are reporting errors for sending attestations to peers. If the clock sync is a little out or we receive attestations before blocks, peers are being too harshly penalized. They can get scored many times per missing block and we typically need these peers on subnets. 


## Proposed Changes

This removes the penalization for missing blocks with attestations. The penalty should be handled when #635 gets built as it will allow us to group attestations per missing block and penalize once.
2020-12-10 00:40:12 +00:00
Age Manning
4f85371ce8 Downgrades a valid log (#2057)
## Issue Addressed

#2046 

## Proposed Changes

The log was originally intended to verify the correct logic and ordering of events when scoring peers. The queued tasks can be structured in such a way that peers can be banned after they are disconnected. Therefore the error log is now downgraded to  debug log.
2020-12-08 10:48:45 +00:00
divma
f3200784b4 More metrics + RPC tweaks (#2041)
## Issue Addressed

NA

## Proposed Changes
This was mostly done to find the reason why LH was dropping peers from Nimbus. It proved to be useful so I think it's worth it. But there is also some functional stuff here
- Add metrics for rpc errors per client, error type and direction
- Add metrics for downscoring events per source type, client and penalty type
- Add metrics for gossip validation results per client for non-accepted messages
- Make the RPC handler return errors and requests/responses in the order we see them
- Allow a small burst for the Ping rate limit, from 1 every 5 seconds to 2 every 10 seconds
- Send rate limiting errors with a particular code and use that same code to identify them. I picked something different to 128 since that is most likely what other clients are using for their own errors
- Remove some unused code in the `PeerAction` and the rpc handler
- Remove the unused variant `RateLimited`. tTis was never produced directly, since the only way to get the request's protocol is via de handler. The handler upon receiving from LH a response with an error (rate limited in this case) emits this event with the missing info (It was always like this, just pointing out that we do downscore rate limiting errors regardless of the change)

Metrics for Nimbus looked like this:
Downscoring events: `increase(libp2p_peer_actions_per_client{client="Nimbus"}[5m])`
![image](https://user-images.githubusercontent.com/26765164/101210880-862bf280-3676-11eb-94c0-399f0bf5aa2e.png)

RPC Errors: `increase(libp2p_rpc_errors_per_client{client="Nimbus"}[5m])`
![image](https://user-images.githubusercontent.com/26765164/101210997-ba071800-3676-11eb-847a-f32405ede002.png)

Unaccepted gossip message: `increase(gossipsub_unaccepted_messages_per_client{client="Nimbus"}[5m])`
![image](https://user-images.githubusercontent.com/26765164/101211124-f470b500-3676-11eb-9459-132ecff058ec.png)
2020-12-08 03:55:50 +00:00
Michael Sproul
c1ec386d18 Pass failed gossip blocks to the slasher (#2047)
## Issue Addressed

Closes #2042

## Proposed Changes

Pass blocks that fail gossip verification to the slasher. Blocks that are successfully verified are not passed immediately, but will be passed as part of full block verification.
2020-12-04 05:03:30 +00:00
realbigsean
fdfb81a74a Server sent events (#1920)
## Issue Addressed

Resolves #1434 (this is the last major feature in the standard spec. There are only a couple of places we may be off-spec due to recent spec changes or ongoing discussion)
Partly addresses #1669
 
## Proposed Changes

- remove the websocket server
- remove the `TeeEventHandler` and `NullEventHandler` 
- add server sent events according to the eth2 API spec

## Additional Info

This is according to the currently unmerged PR here: https://github.com/ethereum/eth2.0-APIs/pull/117


Co-authored-by: realbigsean <seananderson33@gmail.com>
2020-12-04 00:18:58 +00:00
blacktemplar
d8cda2d86e Fix new clippy lints (#2036)
## Issue Addressed

NA

## Proposed Changes

Fixes new clippy lints in the whole project (mainly [manual_strip](https://rust-lang.github.io/rust-clippy/master/index.html#manual_strip) and [unnecessary_lazy_evaluations](https://rust-lang.github.io/rust-clippy/master/index.html#unnecessary_lazy_evaluations)). Furthermore, removes `to_string()` calls on literals when used with the `?`-operator.
2020-12-03 01:10:26 +00:00
divma
8fcd22992c No string in slog (#2017)
## Issue Addressed

Following slog's documentation, this should help a bit with string allocations. I left it run for two days and mem usage is lower. This is of course anecdotal, but shouldn't harm anyway 

## Proposed Changes

remove `String` creation in logs when possible
2020-11-30 10:33:00 +00:00
Age Manning
a567f788bd Upgrade to tokio 0.3 (#1839)
## Description

This PR updates Lighthouse to tokio 0.3. It includes a number of dependency updates and some structural changes as to how we create and spawn tasks.

This also brings with it a number of various improvements:

- Discv5 update
- Libp2p update
- Fix for recompilation issues
- Improved UPnP port mapping handling
- Futures dependency update
- Log downgrade to traces for rejecting peers when we've reached our max



Co-authored-by: blacktemplar <blacktemplar@a1.net>
2020-11-28 05:30:57 +00:00
divma
fc07cc3fdf Sync metrics (#1975)
## Issue Addressed
- Add metrics to keep track of peer counts by sync type
- Add metric to keep track of the number of syncing chains in range

## Proposed Changes
Plugin to the network metrics update interval and update too the counts for peers wrt to their sync status with us

## Additional Info
For the peer counts
- By the way it is implemented the numbers won't always match to the total peer count in the `libp2p` metric.
- Updating the gauge with every change is messy because it requires to be updated on connection (in the `eth2_libp2p` crate, while metrics are defined in the `network` crate) on Goodbye sent (for an `IrrelevantPeer`) either in the `beacon_processor` or the `peer_manager`, and on disconnection. Since this is not a critical metric I think counting once every second is enough. If you think more accuracy is needed we can do it too, but it would be harder to maintain)

ATM those look like this
![image](https://user-images.githubusercontent.com/26765164/100275387-22137b00-2f60-11eb-93b9-94b0f265240c.png)
2020-11-26 05:23:17 +00:00
divma
3b4afc27bf Status race condition (#1967)
## Issue Addressed

Sync stalls due to race conditions between dc notifications and status processing
2020-11-25 02:15:38 +00:00
divma
6f890c398e Sync Bug fixes (#1950)
## Issue Addressed

Two issues related to empty batches
- Chain target's was not being advanced when the batch was successful, empty and the chain didn't have an optimistic batch
- Not switching finalized chains. We now switch finalized chains requiring a minimum work first
2020-11-24 02:11:31 +00:00
Michael Sproul
5828ff1204 Implement slasher (#1567)
This is an implementation of a slasher that lives inside the BN and can be enabled via `lighthouse bn --slasher`.

Features included in this PR:

- [x] Detection of attester slashing conditions (double votes, surrounds existing, surrounded by existing)
- [x] Integration into Lighthouse's attestation verification flow
- [x] Detection of proposer slashing conditions
- [x] Extraction of attestations from blocks as they are verified
- [x] Compression of chunks
- [x] Configurable history length
- [x] Pruning of old attestations and blocks
- [x] More tests

Future work:

* Focus on a slice of history separate from the most recent N epochs (e.g. epochs `current - K` to `current - M`)
* Run out-of-process
* Ingest attestations from the chain without a resync

Design notes are here https://hackmd.io/@sproul/HJSEklmPL
2020-11-23 03:43:22 +00:00
Paul Hauner
65b1cf2af1 Add flag to import all attestations (#1941)
## Issue Addressed

NA

## Proposed Changes

Adds the `--import-all-attestations` flag which tells the `network::AttestationService` to import/aggregate all attestations after verification (instead of only ones for subnets that are relevant to local validators).

This is useful for testing/debugging and also for creating back-up nodes that should be all cached up and ready for any validator.

## Additional Info

NA
2020-11-22 23:58:25 +00:00
divma
d0cbf3111a move sync state to the chains KV (#1940)
## Issue Addressed
we have a log saying we add a peer to a chain, and an another one in case the chain is not syncing. To avoid needing to peer there two (and reduce log entries) simply log the chain's syncing state in the chain's KV
2020-11-22 23:58:23 +00:00
divma
d727e55abe Move some rpc processing to the beacon_processor (#1936)
## Issue Addressed
`BlocksByRange` requests were the main culprit of a series of timeouts to peer's requests in general because they produce build up in the router's processor. Those were moved to the blocking executor but a task is being spawned for each; also not ideal since the amount of resources we give to those is not controlled

## Proposed Changes
- Move `BlocksByRange` and `BlocksByRoots` to the `beacon_processor`. The processor crafts the responses and sends them.
- Move too the processing of `StatusMessage`s from other peers. This is a fast operation but it can also build up and won't scale if we keep it in the router (processing one at the time). These don't need to send an answer, so there is no harm in processing them "later" if that were to happen. Sending responses to status requests is still in the router, so we answer as soon as we see them.
- Some "extras" that are basically clean up:
  - Split the `Worker` logic in sync methods (chain processing and rpc blocks), gossip methods (the majority of methods) and rpc methods (the new ones)
  - Move the `status_message` function previously provided by the router's processor to a more central place since it is used by the router, sync, network_context and beacon_processor
 - Some spelling

## Additional Info
What's left to decide/test more thoroughly is the length of the queues and the priority rules. @paulhauner suggested at some point to put status above attestations, and @AgeManning had described an importance of "protecting gossipsub" so my solution is leaving status requests in the router and RPC methods below attestations. Slashings and Exits are at the end.
2020-11-19 23:33:44 +00:00
Pawan Dhananjay
e47739047d Add additional libp2p tests (#1867)
## Issue Addressed

N/A

## Proposed Changes

Adds tests for the eth2_libp2p crate.
2020-11-19 22:32:09 +00:00
blacktemplar
3408de8151 Avoid string initialization in network metrics and replace by &str where possible (#1898)
## Issue Addressed

NA

## Proposed Changes

Removes most of the temporary string initializations in network metrics and replaces them by directly using `&str`. This further improves on PR https://github.com/sigp/lighthouse/pull/1895.

For the subnet id handling the current approach uses a build script to create a static map. This has the disadvantage that the build script hardcodes the number of subnets. If we want to use more than 64 subnets we need to adjust this in the build script.

## Additional Info

We still have some string initializations for the enum `PeerKind`. To also replace that by `&str` I created a PR in the libp2p dependency: https://github.com/sigp/rust-libp2p/pull/91. Either we wait with merging until this dependency PR is merged (and all conflicts with the newest libp2p version are resolved) or we just merge as is and I will create another PR when the dependency is ready.
2020-11-18 23:31:37 +00:00
divma
398919b5d4 router: drop requests from peers that have dc'd (#1919)
## Issue Addressed

A peer might send a lot of requests that comply to the rate limit and the disconnect, this humongous pr makes sure we don't process them if the peer is not connected
2020-11-17 02:06:21 +00:00
Age Manning
49c4630045 Performance improvement for db reads (#1909)
This PR adds a number of improvements:
- Downgrade a warning log when we ignore blocks for gossipsub processing
- Revert a a correction to improve logging of peer score changes
- Shift syncing DB reads off the core-executor allowing parallel processing of large sync messages
- Correct the timeout logic of RPC chunk sends, giving more time before timing out RPC outbound messages.
2020-11-16 07:28:30 +00:00
divma
eb56140582 Update logs + do not downscore peers if WE time out (#1901)
## Issue Addressed

- RPC Errors were being logged twice: first in the peer manager and then again in the router, so leave just the peer manager's one 
- The "reduce peer count" warn message gets thrown to the user for every missed chunk, so instead print it when the request times out and also do not include there info that is not relevant to the user
- The processor didn't have the service tag so add it
- Impl `KV` for status message
- Do not downscore peers if we are the ones that timed out

Other small improvements
2020-11-16 04:06:14 +00:00
divma
8a16548715 Misc Peer sync info adjustments (#1896)
## Issue Addressed
#1856 

## Proposed Changes
- For clarity, the router's processor now only decides if a peer is compatible and it disconnects it or sends it to sync accordingly. No logic here regarding how useful is the peer. 
- Update peer_sync_info's rules
- Add an `IrrelevantPeer` sync status to account for incompatible peers (maybe this should be "IncompatiblePeer" now that I think about it?) this state is update upon receiving an internal goodbye in the peer manager
- Misc code cleanups
- Reduce the need to create `StatusMessage`s (and thus, `Arc` accesses )
- Add missing calls to update the global sync state

The overall effect should be:
- More peers recognized as Behind, and less as Unknown
- Peers identified as incompatible
2020-11-13 09:00:10 +00:00
Age Manning
c00e6c2c6f Small network adjustments (#1884)
## Issue Addressed

- Asymmetric pings - Currently with symmetric ping intervals, lighthouse nodes race each other to ping often ending in simultaneous ping connections. This shifts the ping interval to be asymmetric based on inbound/outbound connections
- Correct inbound/outbound peer-db registering - It appears we were accounting inbound as outbound and vice versa in the peerdb, this has been corrected
- Improved logging

There is likely more to come - I'll leave this open as we investigate further testnets
2020-11-13 06:06:33 +00:00
Paul Hauner
8772c02fa0 Reduce temp allocations in network metrics (#1895)
## Issue Addressed

Using `heaptrack` I could see that ~75% of Lighthouse temporary allocations are caused by temporary string allocations here.

## Proposed Changes

Reduces temporary `String` allocations when updating metrics in the `network` crate. The solution isn't perfect since we rebuild our caches with each call, but it's a significant improvement.

## Additional Info

NA
2020-11-13 04:19:38 +00:00
blacktemplar
7404f1ce54 Gossipsub scoring (#1668)
## Issue Addressed

#1606 

## Proposed Changes

Uses dynamic gossipsub scoring parameters depending on the number of active validators as specified in https://gist.github.com/blacktemplar/5c1862cb3f0e32a1a7fb0b25e79e6e2c.

## Additional Info

Although the parameters got tested on Medalla, extensive testing using simulations on larger networks is still to be done and we expect that we need to change the parameters, although this might only affect constants within the dynamic parameter framework.
2020-11-12 01:48:28 +00:00
Age Manning
e2ae5010a6 Update libp2p (#1865)
Updates libp2p to the latest version. 

This adds tokio 0.3 support and brings back yamux support. 

This also updates some discv5 configuration parameters for leaner discovery queries
2020-11-06 04:14:14 +00:00
divma
6c0c050fbb Tweak head syncing (#1845)
## Issue Addressed

Fixes head syncing

## Proposed Changes

- Get back to statusing peers after removing chain segments and making the peer manager deal with status according to the Sync status, preventing an old known deadlock
- Also a bug where a chain would get removed if the optimistic batch succeeds being empty

## Additional Info

Tested on Medalla and looking good
2020-11-01 23:37:39 +00:00
divma
9f45ac2f5e More sync edge cases + prettify range (#1834)
## Issue Addressed
Sync edge case when we get an empty optimistic batch that passes validation and is inside the download buffer. Eventually the chain would reach the batch and treat it as an ugly state. 

## Proposed Changes
- Handle the edge case advancing the chain's target + code clarification
- Some largey changes for readability + ergonomics since rust has try ops
- Better handling of bad batch and chain states
2020-10-29 02:29:24 +00:00
Age Manning
7870b81ade Downgrade libp2p (#1817)
## Description

This downgrades the recent libp2p upgrade. 

There were issues with the RPC which prevented syncing of the chain and this upgrade needs to be further investigated.
2020-10-23 09:33:59 +00:00
Age Manning
ae96dab5d2 Increase UPnP logging and decrease batch sizes (#1812)
## Description

This increases the logging of the underlying UPnP tasks to inform the user of UPnP error/success. 

This also decreases the batch syncing size to two epochs per batch.
2020-10-23 03:01:33 +00:00
Age Manning
c49dd94e20 Update to latest libp2p (#1810)
## Description

Updates to the latest libp2p and includes gossipsub updates. 

Of particular note is the limitation of a single topic per gossipsub message.

Co-authored-by: blacktemplar <blacktemplar@a1.net>
2020-10-23 03:01:31 +00:00
divma
668513b67e Sync state adjustments (#1804)
check for advanced peers and the state of the chain wrt the clock slot to decide if a chain is or not synced /transitioning to a head sync. Also a fix that prevented getting the right state while syncing heads
2020-10-22 00:26:06 +00:00
divma
2acf75785c More sync updates (#1791)
## Issue Addressed
#1614 and a couple of sync-stalling problems, the most important is a cyclic dependency between the sync manager and the peer manager
2020-10-20 22:34:18 +00:00
Michael Sproul
703c33bdc7 Fix head tracker concurrency bugs (#1771)
## Issue Addressed

Closes #1557

## Proposed Changes

Modify the pruning algorithm so that it mutates the head-tracker _before_ committing the database transaction to disk, and _only if_ all the heads to be removed are still present in the head-tracker (i.e. no concurrent mutations).

In the process of writing and testing this I also had to make a few other changes:

* Use internal mutability for all `BeaconChainHarness` functions (namely the RNG and the graffiti), in order to enable parallel calls (see testing section below).
* Disable logging in harness tests unless the `test_logger` feature is turned on

And chose to make some clean-ups:

* Delete the `NullMigrator`
* Remove type-based configuration for the migrator in favour of runtime config (simpler, less duplicated code)
* Use the non-blocking migrator unless the blocking migrator is required. In the store tests we need the blocking migrator because some tests make asserts about the state of the DB after the migration has run.
* Rename `validators_keypairs` -> `validator_keypairs` in the `BeaconChainHarness`

## Testing

To confirm that the fix worked, I wrote a test using [Hiatus](https://crates.io/crates/hiatus), which can be found here:

https://github.com/michaelsproul/lighthouse/tree/hiatus-issue-1557

That test can't be merged because it inserts random breakpoints everywhere, but if you check out that branch you can run the test with:

```
$ cd beacon_node/beacon_chain
$ cargo test --release --test parallel_tests --features test_logger
```

It should pass, and the log output should show:

```
WARN Pruning deferred because of a concurrent mutation, message: this is expected only very rarely!
```

## Additional Info

This is a backwards-compatible change with no impact on consensus.
2020-10-19 05:58:39 +00:00
Pawan Dhananjay
97be2ca295 Simulator and attestation service fixes (#1747)
## Issue Addressed

#1729 #1730 

Which issue # does this PR address?

## Proposed Changes

1. Fixes a bug in the simulator where nodes can't find each other due to 0 udp ports in their enr.
2. Fixes bugs in attestation service where we are unsubscribing from a subnet prematurely.

More testing is needed for attestation service fixes.
2020-10-15 07:11:31 +00:00
blacktemplar
8248afa793 Updates the message-id according to the Networking Spec (#1752)
## Proposed Changes

Implement the new message id function (see https://github.com/ethereum/eth2.0-specs/pull/2089) using an additional fast message id function for better performance + caching decompressed data.
2020-10-14 06:51:58 +00:00
Paul Hauner
0e4cc50262
Remove unused deps 2020-10-09 15:58:20 +11:00
Paul Hauner
db3e0578e9
Merge branch 'v0.3.0-staging' into v3-master 2020-10-09 15:27:08 +11:00
Paul Hauner
da44821e39
Clean up obsolete TODOs (#1734)
Squashed commit of the following:

commit f99373cbaec9adb2bdbae3f7e903284327962083
Author: Age Manning <Age@AgeManning.com>
Date:   Mon Oct 5 18:44:09 2020 +1100

    Clean up obsolute TODOs
2020-10-05 21:08:14 +11:00
Paul Hauner
ee7c8a0b7e Update external deps (#1711)
## Issue Addressed

- Resolves #1706 

## Proposed Changes

Updates dependencies across the workspace. Any crate that was not able to be brought to the latest version is listed in #1712.

## Additional Info

NA
2020-10-05 08:22:19 +00:00
Age Manning
240181e840
Upgrade discovery and restructure task execution (#1693)
* Initial rebase

* Remove old code

* Correct release tests

* Rebase commit

* Remove eth2-testnet dep on eth2libp2p

* Remove crates lost in rebase

* Remove unused dep
2020-10-05 18:45:54 +11:00
Age Manning
bcb629564a
Improve error handling in network processing (#1654)
* Improve error handling in network processing

* Cargo fmt

* Cargo fmt

* Improve error handling for prior genesis

* Remove dep
2020-10-05 17:34:56 +11:00
divma
113758a4f5
From panic to crit (#1726)
## Issue Addressed
Downgrade inconsistent chain segment states from `panic` to `crit`. I don't love this solution but since range can always bounce back from any of those, we don't panic.

Co-authored-by: Age Manning <Age@AgeManning.com>
2020-10-05 17:34:49 +11:00
divma
6997776494
Sync fixes (#1716)
## Issue Addressed

chain state inconsistencies

## Proposed Changes
- a batch can be fake-failed by Range if it needs to move a peer to another chain. The peer will still send blocks/ errors / produce timeouts for those  requests, so check when we get a response from the RPC that the request id matches, instead of only the peer, since a re-request can be directed to the same peer.
- if an optimistic batch succeeds, store the attempt to avoid trying it again when quickly switching chains. Also, use it only if ahead of our current target, instead of the segment's start epoch
2020-10-05 17:33:36 +11:00
Paul Hauner
e7eb99cb5e
Use Drop impl to send worker idle message (#1718)
## Issue Addressed

NA

## Proposed Changes

Uses a `Drop` implementation to help ensure that `BeaconProcessor` workers are freed. This will help prevent against regression, if someone happens to add an early return and it will also help in the case of a panic.

## Additional Info

NA
2020-10-05 17:33:25 +11:00
Age Manning
fe07a3c21c
Improve error handling in network processing (#1654)
* Improve error handling in network processing

* Cargo fmt

* Cargo fmt

* Improve error handling for prior genesis

* Remove dep
2020-10-05 17:30:43 +11:00
divma
b1c121b880 From panic to crit (#1726)
## Issue Addressed
Downgrade inconsistent chain segment states from `panic` to `crit`. I don't love this solution but since range can always bounce back from any of those, we don't panic.

Co-authored-by: Age Manning <Age@AgeManning.com>
2020-10-05 04:02:09 +00:00