garage/src/table/schema.rs
Alex b44d3fc796
All checks were successful
continuous-integration/drone/push Build is passing
Abstract database behind generic interface and implement alternative drivers (#322)
- [x] Design interface
- [x] Implement Sled backend
  - [x] Re-implement the SledCountedTree hack ~~on Sled backend~~ on all backends (i.e. over the abstraction)
- [x] Convert Garage code to use generic interface
- [x] Proof-read converted Garage code
- [ ] Test everything well
- [x] Implement sqlite backend
- [x] Implement LMDB backend
- [ ] (Implement Persy backend?)
- [ ] (Implement other backends? (like RocksDB, ...))
- [x] Implement backend choice in config file and garage server module
- [x] Add CLI for converting between DB formats
- Exploit the new interface to put more things in transactions
  - [x] `.updated()` trigger on Garage tables

Fix #284

**Bugs**

- [x] When exporting sqlite, trees iterate empty??
- [x] LMDB doesn't work

**Known issues for various back-ends**

- Sled:
  - Eats all my RAM and also all my disk space
  - `.len()` has to traverse the whole table
  - Is actually quite slow on some operations
  - And is actually pretty bad code...
- Sqlite:
  - Requires a lock to be taken on all operations. The lock is also taken when iterating on a table with `.iter()`, and the lock isn't released until the iterator is dropped. This means that we must be VERY carefull to not do anything else inside a `.iter()` loop or else we will have a deadlock! Most such cases have been eliminated from the Garage codebase, but there might still be some that remain. If your Garage-over-Sqlite seems to hang/freeze, this is the reason.
  - (adapter uses a bunch of unsafe code)
- Heed (LMDB):
  - Not suited for 32-bit machines as it has to map the whole DB in memory.
  - (adpater uses a tiny bit of unsafe code)

**My recommendation:** avoid 32-bit machines and use LMDB as much as possible.

**Converting databases** is actually quite easy. For example from Sled to LMDB:

```bash
cd src/db
cargo run --features cli --bin convert -- -i path/to/garage/meta/db -a sled -o path/to/garage/meta/db.lmdb -b lmdb
```

Then, just add this to your `config.toml`:

```toml
db_engine = "lmdb"
```

Co-authored-by: Alex Auvolat <alex@adnab.me>
Reviewed-on: #322
Co-authored-by: Alex <alex@adnab.me>
Co-committed-by: Alex <alex@adnab.me>
2022-06-08 10:01:44 +02:00

102 lines
2.7 KiB
Rust

use serde::{Deserialize, Serialize};
use garage_db as db;
use garage_util::data::*;
use crate::crdt::Crdt;
/// Trait for field used to partition data
pub trait PartitionKey {
/// Get the key used to partition
fn hash(&self) -> Hash;
}
impl PartitionKey for String {
fn hash(&self) -> Hash {
blake2sum(self.as_bytes())
}
}
/// Values of type FixedBytes32 are assumed to be random,
/// either a hash or a random UUID. This means we can use
/// them directly as an index into the hash table.
impl PartitionKey for FixedBytes32 {
fn hash(&self) -> Hash {
*self
}
}
/// Trait for field used to sort data
pub trait SortKey {
/// Get the key used to sort
fn sort_key(&self) -> &[u8];
}
impl SortKey for String {
fn sort_key(&self) -> &[u8] {
self.as_bytes()
}
}
impl SortKey for FixedBytes32 {
fn sort_key(&self) -> &[u8] {
self.as_slice()
}
}
/// Trait for an entry in a table. It must be sortable and partitionnable.
pub trait Entry<P: PartitionKey, S: SortKey>:
Crdt + PartialEq + Clone + Serialize + for<'de> Deserialize<'de> + Send + Sync
{
/// Get the key used to partition
fn partition_key(&self) -> &P;
/// Get the key used to sort
fn sort_key(&self) -> &S;
/// Is the entry a tombstone? Default implementation always return false
fn is_tombstone(&self) -> bool {
false
}
}
/// Trait for the schema used in a table
pub trait TableSchema: Send + Sync {
/// The name of the table in the database
const TABLE_NAME: &'static str;
/// The partition key used in that table
type P: PartitionKey + Clone + PartialEq + Serialize + for<'de> Deserialize<'de> + Send + Sync;
/// The sort key used int that table
type S: SortKey + Clone + Serialize + for<'de> Deserialize<'de> + Send + Sync;
/// They type for an entry in that table
type E: Entry<Self::P, Self::S>;
/// The type for a filter that can be applied to select entries
/// (e.g. filter out deleted entries)
type Filter: Clone + Serialize + for<'de> Deserialize<'de> + Send + Sync;
// Action to take if not able to decode current version:
// try loading from an older version
/// Try migrating an entry from an older version
fn try_migrate(_bytes: &[u8]) -> Option<Self::E> {
None
}
/// Actions triggered by data changing in a table. If such actions
/// include updates to the local database that should be applied
/// atomically with the item update itself, a db transaction is
/// provided on which these changes should be done.
/// This function can return a DB error but that's all.
fn updated(
&self,
_tx: &mut db::Transaction,
_old: Option<&Self::E>,
_new: Option<&Self::E>,
) -> db::TxOpResult<()> {
Ok(())
}
fn matches_filter(entry: &Self::E, filter: &Self::Filter) -> bool;
}