Skip to content

Commit

Permalink
Add support for deterministic database shards (#2688):
Browse files Browse the repository at this point in the history
This commit, if merged, adds support to allow multiple indepedent nodes to
produce a binary identical shard for a given range of ledgers. The advantage
is that servers can use content-addressable storage, and can more efficiently
retrieve shards by downloading from multiple peers at once and then verifying
the integrity of a shard by cross-checking its checksum with the checksum
other servers report.
  • Loading branch information
p2peer committed Jun 24, 2020
1 parent 95b8f31 commit 4dc08f8
Show file tree
Hide file tree
Showing 9 changed files with 780 additions and 23 deletions.
1 change: 1 addition & 0 deletions Builds/CMake/RippledCore.cmake
Expand Up @@ -508,6 +508,7 @@ target_sources (rippled PRIVATE
src/ripple/nodestore/impl/DatabaseNodeImp.cpp
src/ripple/nodestore/impl/DatabaseRotatingImp.cpp
src/ripple/nodestore/impl/DatabaseShardImp.cpp
src/ripple/nodestore/impl/DeterministicShard.cpp
src/ripple/nodestore/impl/DecodedBlob.cpp
src/ripple/nodestore/impl/DummyScheduler.cpp
src/ripple/nodestore/impl/EncodedBlob.cpp
Expand Down
19 changes: 19 additions & 0 deletions src/ripple/nodestore/Backend.h
Expand Up @@ -58,6 +58,25 @@ class Backend
virtual void
open(bool createIfMissing = true) = 0;

/** Open the backend.
@param createIfMissing Create the database files if necessary.
@param appType Deterministic appType used to create a backend.
@param uid Deterministic uid used to create a backend.
@param salt Deterministic salt used to create a backend.
This allows the caller to catch exceptions.
*/
virtual void
open(
bool createIfMissing,
boost::optional<uint64_t> appType,
boost::optional<uint64_t> uid,
boost::optional<uint64_t> salt)
{
Throw<std::runtime_error>(std::string(
"Deterministic appType/uid/salt not supported by backend " +
getName()));
}

/** Close the backend.
This allows the caller to catch exceptions.
*/
Expand Down
103 changes: 103 additions & 0 deletions src/ripple/nodestore/DeterministicShard.md
@@ -0,0 +1,103 @@
# Deterministic Database Shards

This doc describes the standard way to assemble the database shard. A shard assembled using this approach becomes deterministic i.e. if two independent sides assemble the shard consists of the same ledgers, accounts and transactions, then they will obtain the same shard files `nudb.dat` and `nudb.key`. The approach deals with the `NuDB` database format only, refer to `https://github.com/vinniefalco/NuDB`.


## Headers

Due to NuDB database definition, the following headers are used for database files:

nudb.key:
```
char[8] Type The characters "nudb.key"
uint16 Version Holds the version number
uint64 UID Unique ID generated on creation
uint64 Appnum Application defined constant
uint16 KeySize Key size in bytes
uint64 Salt A random seed
uint64 Pepper The salt hashed
uint16 BlockSize Size of a file block in bytes
uint16 LoadFactor Target fraction in 65536ths
uint8[56] Reserved Zeroes
uint8[] Reserved Zero-pad to block size
```

nudb.dat:
```
char[8] Type The characters "nudb.dat"
uint16 Version Holds the version number
uint64 UID Unique ID generated on creation
uint64 Appnum Application defined constant
uint16 KeySize Key size in bytes
uint8[64] (reserved) Zeroes
```
there all fields are saved using network byte order (most significant byte first).

To make the shard deterministic the following parameters are used as values of header field both for `nudb.key` and `nudb.dat` files.
```
Version 2
UID digest(0)
Appnum digest(2) | 0x5348524400000000 /* 'SHRD' */
KeySize 32
Salt digest(1)
Pepper XXH64(Salt)
BlockSize 0x1000 (4096 bytes)
LoadFactor 0.5 (numeric 0x8000)
```
Note: XXH64() is well-known hash algorithm.

The `digest(i)` mentioned above defined as the follows:

First, RIPEMD160 hash `H` calculated of the following structure
```
uint256 lastHash Hash of last ledger in shard
uint32 index Index of the shard
uint32 firstSeq Sequence number of first ledger in the shard
uint32 lastSeq Sequence number of last ledger in the shard
uint32 version Version of shard, 2 at the present
```
there all 32-bit integers are hashed in network byte order.

Then, `digest(i)` is defined as the following portion of the above hash `H`:
```
digest(0) = H[0] << 56 | H[2] << 48 | ... | H[14] << 0,
digest(1) = H[1] << 56 | H[3] << 48 | ... | H[15] << 0,
digest(2) = H[19] << 24 | H[18] << 16 | ... | H[16] << 0,
```
where `H[i]` denotes `i`-th byte of hash `H`.


## Contents

After deterministic shard is created using the above mentioned headers, it filled with objects. First, all objects of the shard are collected and sorted in according to their hashes. Here the objects are: ledgers, SHAmap tree nodes including accounts and transactions, and final key object with hash 0. Objects are sorted by increasing of their hashes, precisely, by increasing of hex representations of hashes in lexicographic order.

For example, the following is an example of sorted hashes in their hex representation:
```
0000000000000000000000000000000000000000000000000000000000000000
154F29A919B30F50443A241C466691B046677C923EE7905AB97A4DBE8A5C2423
2231553FC01D37A66C61BBEEACBB8C460994493E5659D118E19A8DDBB1444273
272DCBFD8E4D5D786CF11A5444B30FB35435933B5DE6C660AA46E68CF0F5C447
3C062FD9F0BCDCA31ACEBCD8E530D0BDAD1F1D1257B89C435616506A3EE6CB9E
58A0E5AE427CDDC1C7C06448E8C3E4BF718DE036D827881624B20465C3E1334F
...
```

Finally, objects added to the shard one by one in the sorted order from low to high hashes.


## Tests

To perform test to deterministic shards implementation one can enter the following command:
```
rippled --unittest ripple.NodeStore.DatabaseShard
```

The following is the right output of deterministic shards test:
```
ripple.NodeStore.DatabaseShard DatabaseShard deterministic_shard with backend nudb
Iteration 0: RIPEMD160[nudb.key] = 4CFA8985836B549EC99D2E9705707F488DC91E4E
Iteration 0: RIPEMD160[nudb.dat] = 8CC61F503C36339803F8C2FC652C1102DDB889F1
Iteration 1: RIPEMD160[nudb.key] = 4CFA8985836B549EC99D2E9705707F488DC91E4E
Iteration 1: RIPEMD160[nudb.dat] = 8CC61F503C36339803F8C2FC652C1102DDB889F1
```

35 changes: 30 additions & 5 deletions src/ripple/nodestore/backend/NuDBFactory.cpp
Expand Up @@ -38,7 +38,10 @@ namespace NodeStore {
class NuDBBackend : public Backend
{
public:
static constexpr std::size_t currentType = 1;
static constexpr std::uint64_t currentType = 1;
static constexpr std::uint64_t deterministicType = 0x5348524400000000ull;
/* "SHRD" in ASCII */
static constexpr std::uint64_t deterministicMask = 0xFFFFFFFF00000000ull;

beast::Journal const j_;
size_t const keyBytes_;
Expand Down Expand Up @@ -93,7 +96,11 @@ class NuDBBackend : public Backend
}

void
open(bool createIfMissing) override
open(
bool createIfMissing,
boost::optional<uint64_t> appType,
boost::optional<uint64_t> uid,
boost::optional<uint64_t> salt) override
{
using namespace boost::filesystem;
if (db_.is_open())
Expand All @@ -114,8 +121,9 @@ class NuDBBackend : public Backend
dp,
kp,
lp,
currentType,
nudb::make_salt(),
appType.value_or(currentType),
uid.value_or(nudb::make_uid()),
salt.value_or(nudb::make_salt()),
keyBytes_,
nudb::block_size(kp),
0.50,
Expand All @@ -128,10 +136,27 @@ class NuDBBackend : public Backend
db_.open(dp, kp, lp, ec);
if (ec)
Throw<nudb::system_error>(ec);
if (db_.appnum() != currentType)

/** Old value currentType is accepted for appnum in traditional
* databases, new value is used for deterministic shard databases.
* New 64-bit value is constructed from fixed and random parts.
* Fixed part is bounded by bitmask deterministicMask,
* and the value of fixed part is deterministicType.
* Random part depends on the contents of the shard and may be any.
* The contents of appnum field should match either old or new rule.
*/
if (db_.appnum() != appType.value_or(currentType) &&
(appType ||
(db_.appnum() & deterministicMask) != deterministicType))
Throw<std::runtime_error>("nodestore: unknown appnum");
}

void
open(bool createIfMissing) override
{
open(createIfMissing, boost::none, boost::none, boost::none);
}

void
close() override
{
Expand Down

0 comments on commit 4dc08f8

Please sign in to comment.