Simplified distributed block storage with strong consistency, like in Ceph
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

335 lines
10 KiB

#pragma once
#include <sys/types.h>
#include <sys/ioctl.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <unistd.h>
#include <stdint.h>
#include <malloc.h>
#include <linux/fs.h>
#include <liburing.h>
#include <vector>
#include <map>
#include <list>
#include <deque>
#include <set>
#include <functional>
#include "sparsepp/sparsepp/spp.h"
#include "allocator.h"
#include "ringloop.h"
// States are not stored on disk. Instead, they're deduced from the journal
#define ST_IN_FLIGHT 1
#define ST_J_SUBMITTED 2
#define ST_J_WRITTEN 3
#define ST_J_SYNCED 4
#define ST_J_STABLE 5
#define ST_J_MOVE_SYNCED 8
#define ST_D_SUBMITTED 16
#define ST_D_WRITTEN 17
#define ST_D_SYNCED 18
#define ST_D_META_WRITTEN 19
#define ST_D_META_SYNCED 20
#define ST_D_STABLE 21
#define ST_D_META_MOVED 22
#define ST_DEL_WRITTEN 33
#define ST_DEL_SYNCED 34
#define ST_DEL_STABLE 35
#define ST_DEL_MOVED 36
#define ST_CURRENT 48
#define IS_IN_FLIGHT(st) (st == ST_IN_FLIGHT || st == ST_J_SUBMITTED || st == ST_D_SUBMITTED || st == ST_DEL_SUBMITTED)
#define IS_STABLE(st) (st >= ST_J_STABLE && st <= ST_J_MOVE_SYNCED || st >= ST_D_STABLE && st <= ST_D_META_COMMITTED || st >= ST_DEL_STABLE && st <= ST_DEL_MOVED || st == ST_CURRENT)
#define IS_JOURNAL(st) (st >= ST_J_SUBMITTED && st <= ST_J_MOVE_SYNCED)
#define IS_BIG_WRITE(st) (st >= ST_D_SUBMITTED && st <= ST_D_META_COMMITTED)
#define IS_UNSYNCED(st) (st == ST_J_WRITTEN || st >= ST_D_WRITTEN && st <= ST_D_META_WRITTEN || st == ST_DEL_WRITTEN)
// Default object size is 128 KB
#define DEFAULT_ORDER 17
#define MAX_BLOCK_SIZE 128*1024*1024
#define DISK_ALIGNMENT 512
#define STRIPE_NUM(oid) ((oid) >> 4)
#define STRIPE_REPLICA(oid) ((oid) & 0xf)
#define BS_SUBMIT_GET_SQE(sqe, data) \
struct ring_data_t *data = ((ring_data_t*)sqe->user_data)
#define BS_SUBMIT_GET_ONLY_SQE(sqe) \
struct io_uring_sqe *sqe = get_sqe();\
if (!sqe)\
/* Pause until there are more requests available */\
op->wait_for = WAIT_SQE;\
return 0;\
#define BS_SUBMIT_GET_SQE_DECL(sqe) \
sqe = get_sqe();\
if (!sqe)\
/* Pause until there are more requests available */\
op->wait_for = WAIT_SQE;\
return 0;\
class blockstore;
class blockstore_operation;
// 16 bytes per object/stripe id
// stripe includes replica number in 4 least significant bits
struct __attribute__((__packed__)) object_id
uint64_t inode;
uint64_t stripe;
#include "blockstore_journal.h"
inline bool operator == (const object_id & a, const object_id & b)
return a.inode == b.inode && a.stripe == b.stripe;
inline bool operator < (const object_id & a, const object_id & b)
return a.inode < b.inode || a.inode == b.inode && a.stripe < b.stripe;
// 24 bytes per "clean" entry on disk with fixed metadata tables
// FIXME: maybe add crc32's to metadata
struct __attribute__((__packed__)) clean_disk_entry
object_id oid;
uint64_t version;
// 32 = 16 + 16 bytes per "clean" entry in memory (object_id => clean_entry)
struct __attribute__((__packed__)) clean_entry
uint64_t version;
uint64_t location;
// 56 = 24 + 32 bytes per dirty entry in memory (obj_ver_id => dirty_entry)
struct __attribute__((__packed__)) obj_ver_id
object_id oid;
uint64_t version;
inline bool operator < (const obj_ver_id & a, const obj_ver_id & b)
return a.oid < b.oid || a.oid == b.oid && a.version < b.version;
struct __attribute__((__packed__)) dirty_entry
uint32_t state;
uint32_t flags; // unneeded, but present for alignment
uint64_t location; // location in either journal or data
uint32_t offset; // offset within stripe
uint32_t len; // data length
uint64_t journal_sector; // journal sector used for this entry
class oid_hash
size_t operator()(const object_id &s) const
size_t seed = 0;
spp::hash_combine(seed, s.inode);
spp::hash_combine(seed, s.stripe);
return seed;
// - Sync must be submitted after previous writes/deletes (not before!)
// - Reads to the same object must be submitted after previous writes/deletes
// are written (not necessarily synced) in their location. This is because we
// rely on read-modify-write for erasure coding and we must return new data
// to calculate parity for subsequent writes
// - Writes may be submitted in any order, because they don't overlap. Each write
// goes into a new location - either on the journal device or on the data device
// - Stable (stabilize) must be submitted after sync of that object is completed
// It's even OK to return an error to the caller if that object is not synced yet
// - Journal trim may be processed only after all versions are moved to
// the main storage AND after all read operations for older versions complete
// - If an operation can not be submitted because the ring is full
// we should stop submission of other operations. Otherwise some "scatter" reads
// may end up blocked for a long time.
// Otherwise, the submit order is free, that is all operations may be submitted immediately
// In fact, adding a write operation must immediately result in dirty_db being populated
#define OP_READ 1
#define OP_WRITE 2
#define OP_SYNC 3
#define OP_STABLE 4
#define OP_DELETE 5
#define OP_TYPE_MASK 0x7
// Suspend operation until there are more free SQEs
#define WAIT_SQE 1
// Suspend operation until version <wait_detail> of object <oid> is written
#define WAIT_IN_FLIGHT 2
// Suspend operation until there are <wait_detail> bytes of free space in the journal on disk
#define WAIT_JOURNAL 3
// Suspend operation until the next journal sector buffer is free
struct blockstore_operation
// flags contain operation type and possibly other flags
uint64_t flags;
// finish callback
std::function<void (blockstore_operation*)> callback;
// For reads, writes & deletes: oid is the requested object
object_id oid;
// For reads: version=0 -> last stable, version=UINT64_MAX -> last unstable, version=X -> specific version
// For writes & deletes: a new version is assigned automatically
uint64_t version;
// For reads & writes: offset & len are the requested part of the object, buf is the buffer
uint32_t offset;
// For stabilize requests: buf contains <len> obj_ver_id's to stabilize
uint32_t len;
uint8_t *buf;
int retval;
// FIXME: Move internal fields somewhere
friend class blockstore;
friend class blockstore_journal_check_t;
friend void prepare_journal_sector_write(journal_t & journal, io_uring_sqe *sqe, std::function<void(ring_data_t*)> cb);
// Wait status
int wait_for;
uint64_t wait_detail;
int pending_ops;
// Read
std::map<uint64_t, struct iovec> read_vec;
// Sync, write
uint64_t min_used_journal_sector, max_used_journal_sector;
// Write
struct iovec iov_zerofill[3];
// Sync
std::vector<obj_ver_id> sync_big_writes, sync_small_writes;
std::list<blockstore_operation*>::iterator in_progress_ptr;
int sync_state, prev_sync_count;
#include "blockstore_init.h"
#include "blockstore_flush.h"
class blockstore
struct ring_consumer_t ring_consumer;
// Another option is
spp::sparse_hash_map<object_id, clean_entry, oid_hash> clean_db;
std::map<obj_ver_id, dirty_entry> dirty_db;
std::list<blockstore_operation*> submit_queue; // FIXME: funny thing is that vector is better here
std::vector<obj_ver_id> unsynced_big_writes, unsynced_small_writes;
std::list<blockstore_operation*> in_progress_syncs; // ...and probably here, too
uint32_t block_order, block_size;
uint64_t block_count;
allocator *data_alloc;
uint8_t *zero_object;
int meta_fd;
int data_fd;
uint64_t meta_offset, meta_size, meta_area, meta_len;
uint64_t data_offset, data_size, data_len;
struct journal_t journal;
journal_flusher_t *flusher;
ring_loop_t *ringloop;
inline struct io_uring_sqe* get_sqe()
return ringloop->get_sqe();
friend class blockstore_init_meta;
friend class blockstore_init_journal;
friend class blockstore_journal_check_t;
friend class journal_flusher_t;
friend class journal_flusher_co;
void calc_lengths(spp::sparse_hash_map<std::string, std::string> & config);
void open_data(spp::sparse_hash_map<std::string, std::string> & config);
void open_meta(spp::sparse_hash_map<std::string, std::string> & config);
void open_journal(spp::sparse_hash_map<std::string, std::string> & config);
// Asynchronous init
int initialized;
int metadata_buf_size;
blockstore_init_meta* metadata_init_reader;
blockstore_init_journal* journal_init_reader;
void check_wait(blockstore_operation *op);
// Read
int dequeue_read(blockstore_operation *read_op);
int fulfill_read(blockstore_operation *read_op, uint64_t &fulfilled, uint32_t item_start, uint32_t item_end,
uint32_t item_state, uint64_t item_version, uint64_t item_location);
int fulfill_read_push(blockstore_operation *read_op, uint64_t &fulfilled, uint32_t item_start,
uint32_t item_state, uint64_t item_version, uint64_t item_location, uint32_t cur_start, uint32_t cur_end);
void handle_read_event(ring_data_t *data, blockstore_operation *op);
// Write
void enqueue_write(blockstore_operation *op);
int dequeue_write(blockstore_operation *op);
void handle_write_event(ring_data_t *data, blockstore_operation *op);
// Sync
int dequeue_sync(blockstore_operation *op);
void handle_sync_event(ring_data_t *data, blockstore_operation *op);
int continue_sync(blockstore_operation *op);
int ack_sync(blockstore_operation *op);
// Stabilize
int dequeue_stable(blockstore_operation *op);
void handle_stable_event(ring_data_t *data, blockstore_operation *op);
void stabilize_object(object_id oid, uint64_t max_ver);
blockstore(spp::sparse_hash_map<std::string, std::string> & config, ring_loop_t *ringloop);
// Event loop
void loop();
// Returns true when it's safe to destroy the instance. If destroying the instance
// requires to purge some queues, starts that process. Should be called in the event
// loop until it returns true.
bool stop();
// Submission
int enqueue_op(blockstore_operation *op);