Commit Graph

53 Commits

Author SHA1 Message Date
Etan Kissling 4fc1550d0f
add `{.push raises: [].}` to recently modified files (#5908)
Status Nim style mandates `{.push raises: []}.` at start of modules.
Ensure that's the case so that exceptions are properly tracked.

- https://status-im.github.io/nim-style-guide/errors.exceptions.html
- https://github.com/status-im/nim-eth/pull/614#discussion_r1220906149
2024-02-18 01:16:49 +00:00
Etan Kissling 7c53841cd8
Revert "Revert "fix checkpoint block potentially not getting backfilled into DB (#5863)" (#5871)" (#5875)
This reverts commit 1575478b72.
2024-02-09 20:44:54 +01:00
tersec 1575478b72
Revert "fix checkpoint block potentially not getting backfilled into DB (#5863)" (#5871)
This reverts commit 65e6f892de.
2024-02-09 12:49:07 +00:00
Etan Kissling 65e6f892de
fix checkpoint block potentially not getting backfilled into DB (#5863)
When using checkpoint sync, only checkpoint state is available, block is
not downloaded and backfilled later.

`dag.backfill` tracks latest filled `slot`, and latest `parent_root` for
which no block has been synced yet.

In checkpoint sync, this assumption is broken, because there, the start
`dag.backfill.slot` is set based on checkpoint state slot, and the block
is also not available.

However, sync manager in backward mode also requests `dag.backfill.slot`
and `block_clearance` then backfills the checkpoint block once it is
synced. But, there is no guarantee that a peer ever sends us that block.
They could send us all parent blocks and solely omit the checkpoint
block itself. In that situation, we would accept the parent blocks and
advance `dag.backfill`, and subsequently never request the checkpoint
block again, resulting in gap inside blocks DB that is never filled.

To mitigate that, the assumption is restored that `dag.backfill.slot`
is the latest filled `slot`, and `dag.backfill.parent_root` is the next
block that needs to be synced. By setting `slot` to `tail.slot + 1` and
`parent_root` to `tail.root`, we put a fake summary into `dag.backfill`
so that `block_clearance` only proceeds once checkpoint block exists.
2024-02-09 11:20:36 +01:00
Etan Kissling ed8743b986
fix standalone compilation of `trusted_node_sync.nim` (#5861)
#5544 contained a regression that broke standalone compilation of
`trusted_node_sync` as a main module. Fix it, and add to CI.
2024-02-07 19:26:29 +00:00
tersec 6c53dc1e11
automated consensus spec URL updating to v1.4.0-beta.6 (#5804) 2024-01-20 11:19:47 +00:00
Jacek Sieka 62cbdeefc5
verify `genesis_time` more strictly (fixes #1667) (#5694)
Bogus values lead to crashes down the line when timers overflow
2024-01-06 15:26:56 +01:00
tersec 0b9c632b69
fix some `Name` and `XDeclaredButNotUsed` hints (#5670) 2023-12-15 02:00:14 +01:00
tersec 9efb2958ec
automated consensus spec URL updating to v1.4.0-beta.5 (#5647) 2023-12-05 03:34:45 +01:00
tersec 7e3aeaea09
automated consensus spec URL updating to v1.4.0-beta.4 (#5577) 2023-11-08 05:28:03 +00:00
Etan Kissling d289da8cd4
extract trusted node sync trigger into separate function (#5544)
To prepare for calling trusted node sync from the main Nimbus startup
logic, extract the trusted node sync trigger into a separate function.
Further allow passing a pre-opened database, as that will be needed to
check whether trusted node sync needs to be called during regular start.
2023-10-31 01:56:52 +01:00
tersec 4ddd771127
automated consensus spec URL updating to v1.4.0-beta.3 (#5514) 2023-10-19 10:26:38 +00:00
Etan Kissling 297c768816
reduce redundant zero initialization for LC data objects (#5479)
Directly initialize `ForkedLightClientObj` instead of separately first
 setting the `kind` (initializing everything to zero) and then assigning
the forky data after that.
2023-10-04 18:11:45 +02:00
tersec 2895a9a05c
automated consensus spec URL updating to v1.4.0-beta.2 (#5453) 2023-09-21 18:06:51 +00:00
Etan Kissling e7bc41e005
`blck` --> `forkyBlck` when using `withBlck` / `withStateAndBlck` (#5451)
For symmetry with `forkyState` when using `withState`, and to avoid
problems with shadowing of `blck` when using `withBlck` in `template`,
also rename the injected `blck` to `forkyBlck`.

- https://github.com/nim-lang/Nim/issues/22698
2023-09-21 12:49:14 +02:00
tersec 85e1976ac3
automated consensus spec URL updating to v1.4.0-beta.1 (#5280) 2023-08-09 03:58:47 +00:00
Etan Kissling 41b93ae57a
fix code for Nim `v1.6.14` compatibility (#5131)
With `v1.6.14` there is compilation issue in `trusted_node_sync` where
a type is not inferred automatically anymore for a `nil` instance.
Fix it so we can bump the compiler.

See https://github.com/status-im/nimbus-build-system/pull/63
2023-06-28 17:21:43 +00:00
tersec 614202e30d
automated consensus spec URL updating to v1.4.0-beta.0 (#5121) 2023-06-24 15:43:30 +00:00
tersec 788cdb7133
automated v1.4.0-alpha.2 to v1.4.0-alpha.3 consensus spec URL updates (#5065) 2023-06-13 14:03:49 +00:00
Etan Kissling a3229a6b9b
extract helper for validating `UpdatesByRange` consistency (#5053)
Reduce code duplication when checking response of `UpdatesByRange`.
2023-06-12 14:22:32 +02:00
tersec 22208836b1
automated v1.4.0-alpha.1 to v1.4.0-alpha.2 consensus spec URL updates (#5056) 2023-06-10 09:56:54 +00:00
Etan Kissling a0d63c50ee
do not use untrusted `genesis.ssz` with `--trusted-block-root` (#5025)
When using trusted node sync with `--trusted-block-root`, the remote
server is only trusted for data availability, not for correctness.
As a downloaded genesis state cannot be validated for correctness,
require it to be passed via the network metadata `genesis.ssz` file
for `--trusted-block-root` mode. Network metadata is considered trusted
as it is provided by the user and not by the remote server.
Further adds a check for consistent `genesis_time` when using `StateId`
based trusted node sync. This is just a sanity check to avoid spreading
blatantly incorrect data, similar to existing `genesis_validators_root`
checks.
2023-06-09 22:11:14 +02:00
tersec f86febc111
update consensus spec URLs to v1.4.0-alpha.1 (#5027) 2023-06-04 10:32:50 +00:00
tersec 1f535336da
automated v1.3.0 to v1.4.0-alpha.0 consensus spec URL updates (#4996)
* automated v1.3.0 to v1.4.0-alpha.0 consensus spec URL updates

* add copyright year headers
2023-05-26 00:14:28 +02:00
Etan Kissling 508819633a
more consensus-specs v1.3.0 bumps (#4898)
Updates more consensus-specs references to v1.3.0.
This PR only includes updates for sections otherwise unmodified.
2023-05-05 20:38:51 +00:00
Etan Kissling c4e258af2a
check signatures in CP sync backfill with LC (#4858)
When using trusted node sync with light client (`--trusted-block-root`),
the trust assumption on the server is reduced to solely be responsible
for data availability, but not data correctness. This means that we must
check block proposer signatures against the downloaded checkpoint, as
they are not covered by the block root.

Note that this lowers the backfill speed when using LC based CP sync
due to the extra checks, by about 60% for me.
2023-04-26 13:37:27 +02:00
Jacek Sieka 7df75d77fa
move spec constants to spec (#4825)
* move spec constants to spec

quite funny, this removes >_1000_ transitive imports when building
`nimbus_validator_client`.

`genDepend`:

```sh
[arnetheduck@praeceps beacon_chain]$ diff -u
nimbus_validator_client.dot.pre nimbus_validator_client.dot | wc
   1167    3507  103822
```

* imports

* better constants

* oops

* too many blobs

* reorder spec constants in fork order

* cleanup
2023-04-17 17:18:54 +00:00
Etan Kissling cb9e0eed49
allow trusted node sync based on LC trusted block root (#4736)
* allow trusted node sync based on LC trusted block root

Extends `trustedNodeSync` with a new `--trusted-block-root` option that
allows initializing a light client. No `--state-id` must be provided.
The beacon node will then use this light client to obtain the latest
finalized state from the remote server in a trust-minimized fashion.
Note that the provided `--trusted-block-root` should be somewhat recent,
and that security precautions such as comparing the state root against
block explorers is still recommended.

* fix

* workaround for `valueOr` limitations

* reduce magic numbers

* digest len > context len for readability

* move `cstring` conversion to caller

* avoid abbreviations

* `return` codestyle
2023-04-16 06:07:07 +00:00
Etan Kissling ad118cd354
rename `stateFork` > `consensusFork` (#4718)
Just the variable, not yet `lcDataForkAtStateFork` / `atStateFork`.

- Shorten comment in `light_client.nim` to keep line width
- Do not rename `stateFork` mention in `runProposalForkchoiceUpdated`.
- Do not rename `stateFork` in `getStateField(dag.headState, fork)`

Rest is just a mechanical mass replace
2023-03-11 00:35:52 +00:00
tersec a47f0b054e
finish eip4844 to deneb module rename (#4705) 2023-03-09 01:34:17 +01:00
tersec e342fdd97a
rename stateForkAtEpoch to consensusForkAtEpoch (#4627) 2023-02-16 11:32:12 +02:00
tersec aacc8d702d
remove Nim 1.2-compatible `push raise`s and update copyright notice years (#4528) 2023-01-20 14:14:37 +00:00
henridf 64878888bd
Blob storage (#4454)
* Blob storage

* fix indentation

* Fix build (none->Opt.none)

* putBlobs -> putBlobsSidecar

* getBlobs -> getBlobsSidecar

* Check blob correctness when storing a backfill block

* Blobs table: rename and conditionally create

* Check block<->blob match in storeBackfillBlock

* Use when .. toFork() to condition on type

* Check blob viability in block_processor.storeBlock()

* Fix build

* Review feedback
2023-01-09 18:42:10 +00:00
Jacek Sieka 75c7195bfd
Backfill only up to MIN_EPOCHS_FOR_BLOCK_REQUESTS blocks (#4421)
When backfilling, we only need to download blocks that are newer than
MIN_EPOCHS_FOR_BLOCK_REQUESTS - the rest cannot reliably be fetched from
the network and does not have to be provided to others.

This change affects only trusted-node-synced clients - genesis sync
continues to work as before (because it needs to construct a state by
building it from genesis).

Those wishing to complete a backfill should do so with era files
instead.
2022-12-23 08:42:55 +01:00
tersec 7faef7827e
fix EIP4844 withBlck (#4411)
* fix EIP4844 withBlck

* don't raiseAssert by default
2022-12-14 18:30:56 +01:00
zah d30cb8baf1
Support for obtaining deposit snapshots during trustedNodeSync (#4303)
Other changes:

* More optimal search for TTD block.

* Add timeouts to all REST requests during trusted node sync.
  Fixes #4037

* Removed support for storing a deposit snapshot in the network
  metadata.
2022-12-07 12:24:51 +02:00
Etan Kissling 48994f67d3
rename `BlockError` -> `VerifierError` (#4310)
We currently use `BlockError` for both beacon blocks and LC objects.
In light of EIP4844, we will likely also use it for blob sidecars.
To avoid confusion, renaming it to a more generic `VerifierError`,
and update its documentation to be more generic.

To avoid long lines as a followup, also renaming the `block_processor`'s
`BlockProcessingCompleted.completed`->`ProcessingStatus.completed` and
`BlockProcessingCompleted.notCompleted`->`ProcessingStatus.notCompleted`
2022-11-10 17:40:27 +00:00
Jacek Sieka 09ade6d33d
Make trusted node sync era-aware (#4283)
This PR removes a bunch of code to make TNS aware of era files, avoiding
a duplicated backfill when era files are available.

* reuse chaindag for loading backfill state, replacing the TNS homebrew
* fix era block iteration to skip empty slots
* add tests for `can_advance_slots`
2022-11-10 10:44:47 +00:00
Jacek Sieka d839b9d07e
State-only checkpoint state startup (#4251)
Currently, we require genesis and a checkpoint block and state to start
from an arbitrary slot - this PR relaxes this requirement so that we can
start with a state alone.

The current trusted-node-sync algorithm works by first downloading
blocks until we find an epoch aligned non-empty slot, then downloads the
state via slot.

However, current
[proposals](https://github.com/ethereum/beacon-APIs/pull/226) for
checkpointing prefer finalized state as
the main reference - this allows more simple access control and caching
on the server side - in particular, this should help checkpoint-syncing
from sources that have a fast `finalized` state download (like infura
and teku) but are slow when accessing state via slot.

Earlier versions of Nimbus will not be able to read databases created
without a checkpoint block and genesis. In most cases, backfilling makes
the database compatible except where genesis is also missing (custom
networks).

* backfill checkpoint block from libp2p instead of checkpoint source,
when doing trusted node sync
* allow starting the client without genesis / checkpoint block
* perform epoch start slot lookahead when loading tail state, so as to
deal with the case where the epoch start slot does not have a block
* replace `--blockId` with `--state-id` in TNS command line
* when replaying, also look at the parent of the last-known-block (even
if we don't have the parent block data, we can still replay from a
"parent" state) - in particular, this clears the way for implementing
state pruning
* deprecate `--finalized-checkpoint-block` option (no longer needed)
2022-11-02 10:02:38 +00:00
Jacek Sieka af9ec577d0
nicer error message for failed backfill (#4188)
* nicer error message for failed backfill

Many checkpoint sources don't support block download

* RestGenericError -> RestErrorMessage

...and other assorted fixes to bring rest types closer to spec

* fix tests
2022-09-29 23:55:18 +03:00
tersec 02a99543c6
more `withState` `state` -> `forkyState` (#4112) 2022-09-13 14:53:12 +03:00
Jacek Sieka 91a1b4e0c5
better error message on invalid URL (fixes #4023) (#4024) 2022-08-26 15:47:55 +00:00
Zahary Karadjov 7f681af56f
Make it easier to test trustedNodeSync or different networks 2022-08-18 19:48:11 +03:00
Taneli Hukkinen 17bf42316e
Fix typo in log: canoncial->canonical (#3927) 2022-08-01 00:32:08 +00:00
Miran dfd4afc9f2
compatibility with Nim 1.4+ (#3888) 2022-07-29 10:53:42 +00:00
tersec 56cd6e2b45
rm unused imports and fix stylecheck issues (#3784) 2022-06-21 21:01:45 +02:00
tersec 8eb5d5de09
use ZERO_HASH for default(Eth2Digest)/Eth2Digest() in func calls (#3770) 2022-06-18 04:57:37 +00:00
Jacek Sieka b3d80827fb
tns: checkpoint wal periodically while backfilling (#3516)
Witout this, we end up with a massive .wal file that needs to be
checkpointed on first startup (which takes a few minutes) - it's much
more efficient to do smaller checkpoints, it turns out.
2022-03-18 12:32:20 +01:00
Jacek Sieka d0183ccd77
Historical state reindex for trusted node sync (#3452)
When performing trusted node sync, historical access is limited to
states after the checkpoint.

Reindexing restores full historical access by replaying historical
blocks against the state and storing snapshots in the database.

The process can be initiated or resumed at any point in time.
2022-03-11 12:49:47 +00:00
Jacek Sieka 1f89b7f7b9
speed up trusted node backfill (#3371)
With these changes, we can backfill about 400-500 slots/sec, which means
a full backfill of mainnet takes about 2-3h.

However, the CPU is not saturated - neither in server nor in client
meaning that somewhere, there's an artificial inefficiency in the
communication - 16 parallel downloads *should* saturate the CPU.

One plasible cause would be "too many async event loop iterations" per
block request, which would introduce multiple "sleep-like" delays along
the way.

I can push the speed up to 800 slots/sec by increasing parallel
downloads even further, but going after the root cause of the slowness
would be better.

* avoid some unnecessary block copies
* double parallel requests
2022-02-12 12:09:59 +01:00