Commit Graph

17 Commits

Author SHA1 Message Date
Jacek Sieka 05ffe7b2bf
Prune `BlockRef` on finalization (#3513)
Up til now, the block dag has been using `BlockRef`, a structure adapted
for a full DAG, to represent all of chain history. This is a correct and
simple design, but does not exploit the linearity of the chain once
parts of it finalize.

By pruning the in-memory `BlockRef` structure at finalization, we save,
at the time of writing, a cool ~250mb (or 25%:ish) chunk of memory
landing us at a steady state of ~750mb normal memory usage for a
validating node.

Above all though, we prevent memory usage from growing proportionally
with the length of the chain, something that would not be sustainable
over time -  instead, the steady state memory usage is roughly
determined by the validator set size which grows much more slowly. With
these changes, the core should remain sustainable memory-wise post-merge
all the way to withdrawals (when the validator set is expected to grow).

In-memory indices are still used for the "hot" unfinalized portion of
the chain - this ensure that consensus performance remains unchanged.

What changes is that for historical access, we use a db-based linear
slot index which is cache-and-disk-friendly, keeping the cost for
accessing historical data at a similar level as before, achieving the
savings at no percievable cost to functionality or performance.

A nice collateral benefit is the almost-instant startup since we no
longer load any large indicies at dag init.

The cost of this functionality instead can be found in the complexity of
having to deal with two ways of traversing the chain - by `BlockRef` and
by slot.

* use `BlockId` instead of `BlockRef` where finalized / historical data
may be required
* simplify clearance pre-advancement
* remove dag.finalizedBlocks (~50:ish mb)
* remove `getBlockAtSlot` - use `getBlockIdAtSlot` instead
* `parent` and `atSlot` for `BlockId` now require a `ChainDAGRef`
instance, unlike `BlockRef` traversal
* prune `BlockRef` parents on finality (~200:ish mb)
* speed up ChainDAG init by not loading finalized history index
* mess up light client server error handling - this need revisiting :)
2022-03-17 17:42:56 +00:00
Jacek Sieka c64bf045f3
remove StateData (#3507)
One more step on the journey to reduce `BlockRef` usage across the
codebase - this one gets rid of `StateData` whose job was to keep track
of which block was last assigned to a state - these duties have now been
taken over by `latest_block_root`, a fairly recent addition that
computes this block root from state data (at a small cost that should be
insignificant)

99% mechanical change.
2022-03-16 08:20:40 +01:00
Emil 0051af430b Put `application/json` as a higher preference than `application/octet-stream` 2022-01-30 18:50:14 +02:00
Jacek Sieka 0e2b4e39fa
REST JSON support improvements (#3232)
* support downloading blocks / states via JSON in addition to SSZ -
slow, but needed for infura support - SSZ is still used when server
supports it
* use common forked block/state reader in REST API
* fix stack overflows in REST JSON decoder
* fix invalid serialization of `justification_bits` in
`/eth/v1/debug/beacon/states` and `/eth/v2/debug/beacon/states`
* fix REST client to use `/eth/...` instead of `/api/eth/...`, update
"default" urls to expose REST api via `/eth` as well as this is what the
standard says - `/api` was added early on based on an example "base url"
in the spec that has been removed since
* expose Nimbus REST extensions via `/nimbus` in addition to
`/api/nimbus` to stay consistent with `/eth`
* fix invalid state root when reading states via REST
* fix recursive imports in `spec/ssz_codec`
* remove usages of `serialization.useCustomSerialization` - fickle
2022-01-06 08:38:40 +01:00
zah fba1f08a5e
Implement #3129 (Optimized history traversals in the REST API) (#3219)
* Fix REST some rest call signatures and implement a simple API benchmark tool

* Implement #3129 (Optimized history traversals in the REST API)

Other notable changes:

The `updateStateData` procedure in the `blockchain_dag.nim` module is
optimized to not rewind down to the last snapshot state saved in the
database if the supplied input state can be used as a starting point
instead.

* Disallow await in withStateForBlockSlot
2022-01-05 15:49:10 +01:00
tersec b81c06edab
rename Beacon{Block,State}Fork.Merge to Bellatrix; update copyright years (#3240) 2022-01-04 09:45:38 +00:00
Jacek Sieka a086cf01ac
altair fork handling cleanups (#3050)
* fix stack overflow crash in REST/debug/getStateV2
* introduce `ForkyXxx` for generic type matching of `Xxx` across
branches (SomeHashedBeaconState -> ForkyHashedBeaconState et al) -
`Some` is already used for other types of type classes
* consolidate function naming in BeaconChainDB, use some generics
* import `forks.nim` from other spec modules and move `Forked*` helpers
around to resolve circular imports
* remove `ForkedBeaconState`, use `ForkedHashedBeaconState` throughout
(less data shuffling between the types)
* fix several cases of states being stored on stack in tests, causing
random failures on some platforms
* remove reading json support from ncli - this should be ported to the
rest json reading instead (doesn't currently work because stack sizes)
2021-11-05 08:34:34 +01:00
Eugene Kabanov 7c9a6b7170
Improve chronos.Future tracking. (#2988)
* Add `child_id` field.

* Fix json-rpc api call and bump chronos.

* Bump chronos master and fix compilation warnings.

* One more bump of `chronos`.

* add random import

* export rest_utils a bit more

Co-authored-by: Jacek Sieka <jacek@status.im>
2021-10-27 14:01:11 +02:00
Jacek Sieka df3fc9525f
import cleanup (#2997)
* import cleanup

...and remove some unused types

* add random imports

* more imports
2021-10-19 16:09:26 +02:00
Jacek Sieka c40cc6cec1 clean up fork enum and field names
* single naming strategy
* simplify some fork code
* simplify forked block production
2021-10-19 11:06:38 +03:00
Etan Kissling 9ee134324b
allow `withXxx` to access fork-specific fields (#2943)
So far, `withState` and `withBlck` templates could only be used to have
convenience access to fork-agnostic BeaconState and BeaconBlock fields.
This patch:
- injects an additional `stateFork` constant that allows to use
  `when` expressions to also access Altair and Merge-specific fields.
- introduces a `withStateAndBlck` template to support operating on both
  a `BeaconState` and `BeaconBlock` at a time.
- makes sync committee related functions Merge aware.
- changes a couple if-else trees for forks into case statements so that
  forgotten future forks are promoted to compile-time errors.
2021-10-06 20:05:06 +03:00
tersec 2b2846b468
implement forked merge state/block support (#2890)
* implement forked state/block support

* merge support for containsOrphan; import cleanup; 80-column lines

* add merge block header operations and slot sanity fixture

* add epoch state transition tests; implement is_valid_gas_limit(), is_merge_block(), is_execution_enabled(), and compute_timestamp_at_slot()

* implement process_execution_payload() and add merge deposit operations tests

* add merge block sanity tests

* add merge case to syncCommitteeParticipants

* v1.1.0-beta.5 updates

* reduce getTestStates-based memory usage; don't try to REST-serialize ExecutionPayload transactions without underlying support

* add execution payload tests; switch var to let in tests/official/
2021-09-27 14:22:58 +00:00
Eugene Kabanov 7ab1856c04
SSZ encoded responses for REST API calls. (#2851) 2021-09-16 16:32:32 +03:00
Jacek Sieka 01596c45dd
cleanups and fixes (#2827)
* import cleanup
* fix json-rpc exception handlers
* avoid unnecessary presto client import
* introduce ForkedBeaconBlock, some altair logging
* url fixes
2021-08-27 11:00:06 +02:00
Eugene Kabanov 66cb18d69b
Number of REST fixes for Altair. (#2790)
* Fix getForkSchedule call.
Create cache of all configuration endpoints at node startup.
Add prepareJsonResponse() call to create cached responses.
Mark all procedures with `raises`.

* Add getForkSchedule to VC.
Fix getForkSchedule return type for API.
More `raises` annotations.
Fix VC fork_service.nim.

* Use `push raises` instead of inline `raises`.

* Improvements for REST API aggregated attestations and attestations processing.

* Rename eth2_network.sendXXX procedures to eth2_network.broadcastXXX.
Add broadcastBeaconBlock() and broadcastAggregateAndProof().
Fix links to specification in REST API declarations.
Add implementation for v2 getStateV2().
Add validator_duties.sendXXX procedures which not only broadcast data, but also validate it.
Fix JSON-RPC/REST to use new validator_duties.sendXXX procedures instead of own implementations.

* Fix validator_client online nodes count incorrect value.
Fix aggregate and proof attestation could be sent too late.

* Adding timeout for block wait in attestations processing.
Fix compilation errors.

* Attempt to debug aggregate and proofs.

* Fix Beacon AIP to use `sendAttestation`.
Add link comment to produceBlockV2.

* Add debug logs before publish operation for blocks, attestations and aggregated attestations.
Fix attestations publishing issue.

* logging fixes

`indexInCommnittee` already logged in attestation

Co-authored-by: Jacek Sieka <jacek@status.im>
2021-08-23 12:41:48 +02:00
Jacek Sieka 7a622e8505
rework spec imports (#2779)
The spec imports are a mess to work with, so this branch cleans them up
a bit to ensure that we avoid generic sandwitches and that importing
stuff generally becomes easier.

* reexport crypto/digest/presets because these are part of the public
symbol set of the rest of the spec types
* don't export `merge` types from `base` - this causes circular deps
* fix circular deps in `ssz/spec_types` - this is the first step in
disentangling ssz from spec
* be explicit about phase0 vs altair - longer term, `altair` will become
the "natural" type set, then merge and so on, so no point in giving
`phase0` special preferential treatment
2021-08-12 13:08:20 +00:00
Jacek Sieka 3d7bee8502
REST API client, JSON-RPC cleanups (#2756)
This refactoring puts the JSON-RPC and REST APIs on more equal footing
by renaming and moving things around, creating a separation between
client and server, and documenting what they are - the aim is to have a
simple-to-use base to start from when developing API clients, as well as
make it easier to navigate the code when looking for the legacy JSON-RPC
interface vs the new REST API.

* move REST client, serialization and supporting types to spec/eth2_apis
* REST stuff now starts with `rest_`, JSON-RPC stuff starts with `rpc_`,
more or less
* simplify imports such that there's a simple module to import for both
server and client
* map REST type and proc names to yaml spec more closely - in
particular, reuse operation and type names in `rest_types` to make
comparisons against spec more easy
* cleaner separation between client and server modules - modules common
between server and client such as `rest_types` and serialization move to
the spec folder - this allows the client to be built with less knowledge
about server internals
2021-08-03 17:17:11 +02:00