Merge pull request #960 from hashicorp/update-bolt
Updating BoltDb to 1.2
This commit is contained in:
Diptanu Choudhury
commit
c28c89d5c0
|
|
@ -1,6 +1,7 @@
|
|||
{
|
||||
"ImportPath": "github.com/hashicorp/nomad",
|
||||
"GoVersion": "go1.6",
|
||||
"GodepVersion": "v60",
|
||||
"Deps": [
|
||||
{
|
||||
"ImportPath": "github.com/StackExchange/wmi",
|
||||
|
|
@ -125,8 +126,8 @@
|
|||
},
|
||||
{
|
||||
"ImportPath": "github.com/boltdb/bolt",
|
||||
"Comment": "v1.1.0-32-g827f56d",
|
||||
"Rev": "827f56dfb2091be2edb284ffb6113c198f27b033"
|
||||
"Comment": "v1.2.0",
|
||||
"Rev": "c6ba97b89e0454fec9aa92e1d33a4e2c5fc1f631"
|
||||
},
|
||||
{
|
||||
"ImportPath": "github.com/coreos/go-systemd/dbus",
|
||||
|
|
|
|||
|
|
@ -1,54 +1,18 @@
|
|||
TEST=.
|
||||
BENCH=.
|
||||
COVERPROFILE=/tmp/c.out
|
||||
BRANCH=`git rev-parse --abbrev-ref HEAD`
|
||||
COMMIT=`git rev-parse --short HEAD`
|
||||
GOLDFLAGS="-X main.branch $(BRANCH) -X main.commit $(COMMIT)"
|
||||
|
||||
default: build
|
||||
|
||||
bench:
|
||||
go test -v -test.run=NOTHINCONTAINSTHIS -test.bench=$(BENCH)
|
||||
|
||||
# http://cloc.sourceforge.net/
|
||||
cloc:
|
||||
@cloc --not-match-f='Makefile|_test.go' .
|
||||
|
||||
cover: fmt
|
||||
go test -coverprofile=$(COVERPROFILE) -test.run=$(TEST) $(COVERFLAG) .
|
||||
go tool cover -html=$(COVERPROFILE)
|
||||
rm $(COVERPROFILE)
|
||||
|
||||
cpuprofile: fmt
|
||||
@go test -c
|
||||
@./bolt.test -test.v -test.run=$(TEST) -test.cpuprofile cpu.prof
|
||||
race:
|
||||
@go test -v -race -test.run="TestSimulate_(100op|1000op)"
|
||||
|
||||
# go get github.com/kisielk/errcheck
|
||||
errcheck:
|
||||
@echo "=== errcheck ==="
|
||||
@errcheck github.com/boltdb/bolt
|
||||
@errcheck -ignorepkg=bytes -ignore=os:Remove github.com/boltdb/bolt
|
||||
|
||||
fmt:
|
||||
@go fmt ./...
|
||||
test:
|
||||
@go test -v -cover .
|
||||
@go test -v ./cmd/bolt
|
||||
|
||||
get:
|
||||
@go get -d ./...
|
||||
|
||||
build: get
|
||||
@mkdir -p bin
|
||||
@go build -ldflags=$(GOLDFLAGS) -a -o bin/bolt ./cmd/bolt
|
||||
|
||||
test: fmt
|
||||
@go get github.com/stretchr/testify/assert
|
||||
@echo "=== TESTS ==="
|
||||
@go test -v -cover -test.run=$(TEST)
|
||||
@echo ""
|
||||
@echo ""
|
||||
@echo "=== CLI ==="
|
||||
@go test -v -test.run=$(TEST) ./cmd/bolt
|
||||
@echo ""
|
||||
@echo ""
|
||||
@echo "=== RACE DETECTOR ==="
|
||||
@go test -v -race -test.run="TestSimulate_(100op|1000op)"
|
||||
|
||||
.PHONY: bench cloc cover cpuprofile fmt memprofile test
|
||||
.PHONY: fmt test
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
Bolt [](https://drone.io/github.com/boltdb/bolt/latest) [](https://coveralls.io/r/boltdb/bolt?branch=master) [](https://godoc.org/github.com/boltdb/bolt) 
|
||||
Bolt [](https://drone.io/github.com/boltdb/bolt/latest) [](https://coveralls.io/r/boltdb/bolt?branch=master) [](https://godoc.org/github.com/boltdb/bolt) 
|
||||
====
|
||||
|
||||
Bolt is a pure Go key/value store inspired by [Howard Chu's][hyc_symas]
|
||||
|
|
@ -13,7 +13,6 @@ and setting values. That's it.
|
|||
[hyc_symas]: https://twitter.com/hyc_symas
|
||||
[lmdb]: http://symas.com/mdb/
|
||||
|
||||
|
||||
## Project Status
|
||||
|
||||
Bolt is stable and the API is fixed. Full unit test coverage and randomized
|
||||
|
|
@ -22,6 +21,36 @@ Bolt is currently in high-load production environments serving databases as
|
|||
large as 1TB. Many companies such as Shopify and Heroku use Bolt-backed
|
||||
services every day.
|
||||
|
||||
## Table of Contents
|
||||
|
||||
- [Getting Started](#getting-started)
|
||||
- [Installing](#installing)
|
||||
- [Opening a database](#opening-a-database)
|
||||
- [Transactions](#transactions)
|
||||
- [Read-write transactions](#read-write-transactions)
|
||||
- [Read-only transactions](#read-only-transactions)
|
||||
- [Batch read-write transactions](#batch-read-write-transactions)
|
||||
- [Managing transactions manually](#managing-transactions-manually)
|
||||
- [Using buckets](#using-buckets)
|
||||
- [Using key/value pairs](#using-keyvalue-pairs)
|
||||
- [Autoincrementing integer for the bucket](#autoincrementing-integer-for-the-bucket)
|
||||
- [Iterating over keys](#iterating-over-keys)
|
||||
- [Prefix scans](#prefix-scans)
|
||||
- [Range scans](#range-scans)
|
||||
- [ForEach()](#foreach)
|
||||
- [Nested buckets](#nested-buckets)
|
||||
- [Database backups](#database-backups)
|
||||
- [Statistics](#statistics)
|
||||
- [Read-Only Mode](#read-only-mode)
|
||||
- [Mobile Use (iOS/Android)](#mobile-use-iosandroid)
|
||||
- [Resources](#resources)
|
||||
- [Comparison with other databases](#comparison-with-other-databases)
|
||||
- [Postgres, MySQL, & other relational databases](#postgres-mysql--other-relational-databases)
|
||||
- [LevelDB, RocksDB](#leveldb-rocksdb)
|
||||
- [LMDB](#lmdb)
|
||||
- [Caveats & Limitations](#caveats--limitations)
|
||||
- [Reading the Source](#reading-the-source)
|
||||
- [Other Projects Using Bolt](#other-projects-using-bolt)
|
||||
|
||||
## Getting Started
|
||||
|
||||
|
|
@ -309,7 +338,6 @@ type User struct {
|
|||
ID int
|
||||
...
|
||||
}
|
||||
|
||||
```
|
||||
|
||||
### Iterating over keys
|
||||
|
|
@ -320,7 +348,9 @@ iteration over these keys extremely fast. To iterate over keys we'll use a
|
|||
|
||||
```go
|
||||
db.View(func(tx *bolt.Tx) error {
|
||||
// Assume bucket exists and has keys
|
||||
b := tx.Bucket([]byte("MyBucket"))
|
||||
|
||||
c := b.Cursor()
|
||||
|
||||
for k, v := c.First(); k != nil; k, v = c.Next() {
|
||||
|
|
@ -361,6 +391,7 @@ To iterate over a key prefix, you can combine `Seek()` and `bytes.HasPrefix()`:
|
|||
|
||||
```go
|
||||
db.View(func(tx *bolt.Tx) error {
|
||||
// Assume bucket exists and has keys
|
||||
c := tx.Bucket([]byte("MyBucket")).Cursor()
|
||||
|
||||
prefix := []byte("1234")
|
||||
|
|
@ -380,7 +411,7 @@ date range like this:
|
|||
|
||||
```go
|
||||
db.View(func(tx *bolt.Tx) error {
|
||||
// Assume our events bucket has RFC3339 encoded time keys.
|
||||
// Assume our events bucket exists and has RFC3339 encoded time keys.
|
||||
c := tx.Bucket([]byte("Events")).Cursor()
|
||||
|
||||
// Our time range spans the 90's decade.
|
||||
|
|
@ -404,7 +435,9 @@ all the keys in a bucket:
|
|||
|
||||
```go
|
||||
db.View(func(tx *bolt.Tx) error {
|
||||
// Assume bucket exists and has keys
|
||||
b := tx.Bucket([]byte("MyBucket"))
|
||||
|
||||
b.ForEach(func(k, v []byte) error {
|
||||
fmt.Printf("key=%s, value=%s\n", k, v)
|
||||
return nil
|
||||
|
|
@ -517,6 +550,84 @@ if err != nil {
|
|||
}
|
||||
```
|
||||
|
||||
### Mobile Use (iOS/Android)
|
||||
|
||||
Bolt is able to run on mobile devices by leveraging the binding feature of the
|
||||
[gomobile](https://github.com/golang/mobile) tool. Create a struct that will
|
||||
contain your database logic and a reference to a `*bolt.DB` with a initializing
|
||||
contstructor that takes in a filepath where the database file will be stored.
|
||||
Neither Android nor iOS require extra permissions or cleanup from using this method.
|
||||
|
||||
```go
|
||||
func NewBoltDB(filepath string) *BoltDB {
|
||||
db, err := bolt.Open(filepath+"/demo.db", 0600, nil)
|
||||
if err != nil {
|
||||
log.Fatal(err)
|
||||
}
|
||||
|
||||
return &BoltDB{db}
|
||||
}
|
||||
|
||||
type BoltDB struct {
|
||||
db *bolt.DB
|
||||
...
|
||||
}
|
||||
|
||||
func (b *BoltDB) Path() string {
|
||||
return b.db.Path()
|
||||
}
|
||||
|
||||
func (b *BoltDB) Close() {
|
||||
b.db.Close()
|
||||
}
|
||||
```
|
||||
|
||||
Database logic should be defined as methods on this wrapper struct.
|
||||
|
||||
To initialize this struct from the native language (both platforms now sync
|
||||
their local storage to the cloud. These snippets disable that functionality for the
|
||||
database file):
|
||||
|
||||
#### Android
|
||||
|
||||
```java
|
||||
String path;
|
||||
if (android.os.Build.VERSION.SDK_INT >=android.os.Build.VERSION_CODES.LOLLIPOP){
|
||||
path = getNoBackupFilesDir().getAbsolutePath();
|
||||
} else{
|
||||
path = getFilesDir().getAbsolutePath();
|
||||
}
|
||||
Boltmobiledemo.BoltDB boltDB = Boltmobiledemo.NewBoltDB(path)
|
||||
```
|
||||
|
||||
#### iOS
|
||||
|
||||
```objc
|
||||
- (void)demo {
|
||||
NSString* path = [NSSearchPathForDirectoriesInDomains(NSLibraryDirectory,
|
||||
NSUserDomainMask,
|
||||
YES) objectAtIndex:0];
|
||||
GoBoltmobiledemoBoltDB * demo = GoBoltmobiledemoNewBoltDB(path);
|
||||
[self addSkipBackupAttributeToItemAtPath:demo.path];
|
||||
//Some DB Logic would go here
|
||||
[demo close];
|
||||
}
|
||||
|
||||
- (BOOL)addSkipBackupAttributeToItemAtPath:(NSString *) filePathString
|
||||
{
|
||||
NSURL* URL= [NSURL fileURLWithPath: filePathString];
|
||||
assert([[NSFileManager defaultManager] fileExistsAtPath: [URL path]]);
|
||||
|
||||
NSError *error = nil;
|
||||
BOOL success = [URL setResourceValue: [NSNumber numberWithBool: YES]
|
||||
forKey: NSURLIsExcludedFromBackupKey error: &error];
|
||||
if(!success){
|
||||
NSLog(@"Error excluding %@ from backup %@", [URL lastPathComponent], error);
|
||||
}
|
||||
return success;
|
||||
}
|
||||
|
||||
```
|
||||
|
||||
## Resources
|
||||
|
||||
|
|
@ -724,5 +835,10 @@ Below is a list of public, open source projects that use Bolt:
|
|||
backed by boltdb.
|
||||
* [buckets](https://github.com/joyrexus/buckets) - a bolt wrapper streamlining
|
||||
simple tx and key scans.
|
||||
* [mbuckets](https://github.com/abhigupta912/mbuckets) - A Bolt wrapper that allows easy operations on multi level (nested) buckets.
|
||||
* [Request Baskets](https://github.com/darklynx/request-baskets) - A web service to collect arbitrary HTTP requests and inspect them via REST API or simple web UI, similar to [RequestBin](http://requestb.in/) service
|
||||
* [Go Report Card](https://goreportcard.com/) - Go code quality report cards as a (free and open source) service.
|
||||
* [Boltdb Boilerplate](https://github.com/bobintornado/boltdb-boilerplate) - Boilerplate wrapper around bolt aiming to make simple calls one-liners.
|
||||
* [lru](https://github.com/crowdriff/lru) - Easy to use Bolt-backed Least-Recently-Used (LRU) read-through cache with chainable remote stores.
|
||||
|
||||
If you are using Bolt in a project please send a pull request to add it to the list.
|
||||
|
|
|
|||
|
|
@ -0,0 +1,18 @@
|
|||
version: "{build}"
|
||||
|
||||
os: Windows Server 2012 R2
|
||||
|
||||
clone_folder: c:\gopath\src\github.com\boltdb\bolt
|
||||
|
||||
environment:
|
||||
GOPATH: c:\gopath
|
||||
|
||||
install:
|
||||
- echo %PATH%
|
||||
- echo %GOPATH%
|
||||
- go version
|
||||
- go env
|
||||
- go get -v -t ./...
|
||||
|
||||
build_script:
|
||||
- go test -v ./...
|
||||
|
|
@ -1,138 +0,0 @@
|
|||
package bolt
|
||||
|
||||
import (
|
||||
"errors"
|
||||
"fmt"
|
||||
"sync"
|
||||
"time"
|
||||
)
|
||||
|
||||
// Batch calls fn as part of a batch. It behaves similar to Update,
|
||||
// except:
|
||||
//
|
||||
// 1. concurrent Batch calls can be combined into a single Bolt
|
||||
// transaction.
|
||||
//
|
||||
// 2. the function passed to Batch may be called multiple times,
|
||||
// regardless of whether it returns error or not.
|
||||
//
|
||||
// This means that Batch function side effects must be idempotent and
|
||||
// take permanent effect only after a successful return is seen in
|
||||
// caller.
|
||||
//
|
||||
// The maximum batch size and delay can be adjusted with DB.MaxBatchSize
|
||||
// and DB.MaxBatchDelay, respectively.
|
||||
//
|
||||
// Batch is only useful when there are multiple goroutines calling it.
|
||||
func (db *DB) Batch(fn func(*Tx) error) error {
|
||||
errCh := make(chan error, 1)
|
||||
|
||||
db.batchMu.Lock()
|
||||
if (db.batch == nil) || (db.batch != nil && len(db.batch.calls) >= db.MaxBatchSize) {
|
||||
// There is no existing batch, or the existing batch is full; start a new one.
|
||||
db.batch = &batch{
|
||||
db: db,
|
||||
}
|
||||
db.batch.timer = time.AfterFunc(db.MaxBatchDelay, db.batch.trigger)
|
||||
}
|
||||
db.batch.calls = append(db.batch.calls, call{fn: fn, err: errCh})
|
||||
if len(db.batch.calls) >= db.MaxBatchSize {
|
||||
// wake up batch, it's ready to run
|
||||
go db.batch.trigger()
|
||||
}
|
||||
db.batchMu.Unlock()
|
||||
|
||||
err := <-errCh
|
||||
if err == trySolo {
|
||||
err = db.Update(fn)
|
||||
}
|
||||
return err
|
||||
}
|
||||
|
||||
type call struct {
|
||||
fn func(*Tx) error
|
||||
err chan<- error
|
||||
}
|
||||
|
||||
type batch struct {
|
||||
db *DB
|
||||
timer *time.Timer
|
||||
start sync.Once
|
||||
calls []call
|
||||
}
|
||||
|
||||
// trigger runs the batch if it hasn't already been run.
|
||||
func (b *batch) trigger() {
|
||||
b.start.Do(b.run)
|
||||
}
|
||||
|
||||
// run performs the transactions in the batch and communicates results
|
||||
// back to DB.Batch.
|
||||
func (b *batch) run() {
|
||||
b.db.batchMu.Lock()
|
||||
b.timer.Stop()
|
||||
// Make sure no new work is added to this batch, but don't break
|
||||
// other batches.
|
||||
if b.db.batch == b {
|
||||
b.db.batch = nil
|
||||
}
|
||||
b.db.batchMu.Unlock()
|
||||
|
||||
retry:
|
||||
for len(b.calls) > 0 {
|
||||
var failIdx = -1
|
||||
err := b.db.Update(func(tx *Tx) error {
|
||||
for i, c := range b.calls {
|
||||
if err := safelyCall(c.fn, tx); err != nil {
|
||||
failIdx = i
|
||||
return err
|
||||
}
|
||||
}
|
||||
return nil
|
||||
})
|
||||
|
||||
if failIdx >= 0 {
|
||||
// take the failing transaction out of the batch. it's
|
||||
// safe to shorten b.calls here because db.batch no longer
|
||||
// points to us, and we hold the mutex anyway.
|
||||
c := b.calls[failIdx]
|
||||
b.calls[failIdx], b.calls = b.calls[len(b.calls)-1], b.calls[:len(b.calls)-1]
|
||||
// tell the submitter re-run it solo, continue with the rest of the batch
|
||||
c.err <- trySolo
|
||||
continue retry
|
||||
}
|
||||
|
||||
// pass success, or bolt internal errors, to all callers
|
||||
for _, c := range b.calls {
|
||||
if c.err != nil {
|
||||
c.err <- err
|
||||
}
|
||||
}
|
||||
break retry
|
||||
}
|
||||
}
|
||||
|
||||
// trySolo is a special sentinel error value used for signaling that a
|
||||
// transaction function should be re-run. It should never be seen by
|
||||
// callers.
|
||||
var trySolo = errors.New("batch function returned an error and should be re-run solo")
|
||||
|
||||
type panicked struct {
|
||||
reason interface{}
|
||||
}
|
||||
|
||||
func (p panicked) Error() string {
|
||||
if err, ok := p.reason.(error); ok {
|
||||
return err.Error()
|
||||
}
|
||||
return fmt.Sprintf("panic: %v", p.reason)
|
||||
}
|
||||
|
||||
func safelyCall(fn func(*Tx) error, tx *Tx) (err error) {
|
||||
defer func() {
|
||||
if p := recover(); p != nil {
|
||||
err = panicked{p}
|
||||
}
|
||||
}()
|
||||
return fn(tx)
|
||||
}
|
||||
|
|
@ -0,0 +1,9 @@
|
|||
// +build ppc
|
||||
|
||||
package bolt
|
||||
|
||||
// maxMapSize represents the largest mmap size supported by Bolt.
|
||||
const maxMapSize = 0x7FFFFFFF // 2GB
|
||||
|
||||
// maxAllocSize is the size used when creating array pointers.
|
||||
const maxAllocSize = 0xFFFFFFF
|
||||
|
|
@ -0,0 +1,9 @@
|
|||
// +build ppc64
|
||||
|
||||
package bolt
|
||||
|
||||
// maxMapSize represents the largest mmap size supported by Bolt.
|
||||
const maxMapSize = 0xFFFFFFFFFFFF // 256TB
|
||||
|
||||
// maxAllocSize is the size used when creating array pointers.
|
||||
const maxAllocSize = 0x7FFFFFFF
|
||||
|
|
@ -11,7 +11,7 @@ import (
|
|||
)
|
||||
|
||||
// flock acquires an advisory lock on a file descriptor.
|
||||
func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
||||
func flock(db *DB, mode os.FileMode, exclusive bool, timeout time.Duration) error {
|
||||
var t time.Time
|
||||
for {
|
||||
// If we're beyond our timeout then return an error.
|
||||
|
|
@ -27,7 +27,7 @@ func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
|||
}
|
||||
|
||||
// Otherwise attempt to obtain an exclusive lock.
|
||||
err := syscall.Flock(int(f.Fd()), flag|syscall.LOCK_NB)
|
||||
err := syscall.Flock(int(db.file.Fd()), flag|syscall.LOCK_NB)
|
||||
if err == nil {
|
||||
return nil
|
||||
} else if err != syscall.EWOULDBLOCK {
|
||||
|
|
@ -40,23 +40,12 @@ func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
|||
}
|
||||
|
||||
// funlock releases an advisory lock on a file descriptor.
|
||||
func funlock(f *os.File) error {
|
||||
return syscall.Flock(int(f.Fd()), syscall.LOCK_UN)
|
||||
func funlock(db *DB) error {
|
||||
return syscall.Flock(int(db.file.Fd()), syscall.LOCK_UN)
|
||||
}
|
||||
|
||||
// mmap memory maps a DB's data file.
|
||||
func mmap(db *DB, sz int) error {
|
||||
// Truncate and fsync to ensure file size metadata is flushed.
|
||||
// https://github.com/boltdb/bolt/issues/284
|
||||
if !db.NoGrowSync && !db.readOnly {
|
||||
if err := db.file.Truncate(int64(sz)); err != nil {
|
||||
return fmt.Errorf("file resize error: %s", err)
|
||||
}
|
||||
if err := db.file.Sync(); err != nil {
|
||||
return fmt.Errorf("file sync error: %s", err)
|
||||
}
|
||||
}
|
||||
|
||||
// Map the data file to memory.
|
||||
b, err := syscall.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
|
||||
if err != nil {
|
||||
|
|
|
|||
|
|
@ -11,7 +11,7 @@ import (
|
|||
)
|
||||
|
||||
// flock acquires an advisory lock on a file descriptor.
|
||||
func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
||||
func flock(db *DB, mode os.FileMode, exclusive bool, timeout time.Duration) error {
|
||||
var t time.Time
|
||||
for {
|
||||
// If we're beyond our timeout then return an error.
|
||||
|
|
@ -32,7 +32,7 @@ func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
|||
} else {
|
||||
lock.Type = syscall.F_RDLCK
|
||||
}
|
||||
err := syscall.FcntlFlock(f.Fd(), syscall.F_SETLK, &lock)
|
||||
err := syscall.FcntlFlock(db.file.Fd(), syscall.F_SETLK, &lock)
|
||||
if err == nil {
|
||||
return nil
|
||||
} else if err != syscall.EAGAIN {
|
||||
|
|
@ -45,28 +45,17 @@ func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
|||
}
|
||||
|
||||
// funlock releases an advisory lock on a file descriptor.
|
||||
func funlock(f *os.File) error {
|
||||
func funlock(db *DB) error {
|
||||
var lock syscall.Flock_t
|
||||
lock.Start = 0
|
||||
lock.Len = 0
|
||||
lock.Type = syscall.F_UNLCK
|
||||
lock.Whence = 0
|
||||
return syscall.FcntlFlock(uintptr(f.Fd()), syscall.F_SETLK, &lock)
|
||||
return syscall.FcntlFlock(uintptr(db.file.Fd()), syscall.F_SETLK, &lock)
|
||||
}
|
||||
|
||||
// mmap memory maps a DB's data file.
|
||||
func mmap(db *DB, sz int) error {
|
||||
// Truncate and fsync to ensure file size metadata is flushed.
|
||||
// https://github.com/boltdb/bolt/issues/284
|
||||
if !db.NoGrowSync && !db.readOnly {
|
||||
if err := db.file.Truncate(int64(sz)); err != nil {
|
||||
return fmt.Errorf("file resize error: %s", err)
|
||||
}
|
||||
if err := db.file.Sync(); err != nil {
|
||||
return fmt.Errorf("file sync error: %s", err)
|
||||
}
|
||||
}
|
||||
|
||||
// Map the data file to memory.
|
||||
b, err := unix.Mmap(int(db.file.Fd()), 0, sz, syscall.PROT_READ, syscall.MAP_SHARED|db.MmapFlags)
|
||||
if err != nil {
|
||||
|
|
|
|||
|
|
@ -16,6 +16,8 @@ var (
|
|||
)
|
||||
|
||||
const (
|
||||
lockExt = ".lock"
|
||||
|
||||
// see https://msdn.microsoft.com/en-us/library/windows/desktop/aa365203(v=vs.85).aspx
|
||||
flagLockExclusive = 2
|
||||
flagLockFailImmediately = 1
|
||||
|
|
@ -46,7 +48,16 @@ func fdatasync(db *DB) error {
|
|||
}
|
||||
|
||||
// flock acquires an advisory lock on a file descriptor.
|
||||
func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
||||
func flock(db *DB, mode os.FileMode, exclusive bool, timeout time.Duration) error {
|
||||
// Create a separate lock file on windows because a process
|
||||
// cannot share an exclusive lock on the same file. This is
|
||||
// needed during Tx.WriteTo().
|
||||
f, err := os.OpenFile(db.path+lockExt, os.O_CREATE, mode)
|
||||
if err != nil {
|
||||
return err
|
||||
}
|
||||
db.lockfile = f
|
||||
|
||||
var t time.Time
|
||||
for {
|
||||
// If we're beyond our timeout then return an error.
|
||||
|
|
@ -62,7 +73,7 @@ func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
|||
flag |= flagLockExclusive
|
||||
}
|
||||
|
||||
err := lockFileEx(syscall.Handle(f.Fd()), flag, 0, 1, 0, &syscall.Overlapped{})
|
||||
err := lockFileEx(syscall.Handle(db.lockfile.Fd()), flag, 0, 1, 0, &syscall.Overlapped{})
|
||||
if err == nil {
|
||||
return nil
|
||||
} else if err != errLockViolation {
|
||||
|
|
@ -75,8 +86,11 @@ func flock(f *os.File, exclusive bool, timeout time.Duration) error {
|
|||
}
|
||||
|
||||
// funlock releases an advisory lock on a file descriptor.
|
||||
func funlock(f *os.File) error {
|
||||
return unlockFileEx(syscall.Handle(f.Fd()), 0, 1, 0, &syscall.Overlapped{})
|
||||
func funlock(db *DB) error {
|
||||
err := unlockFileEx(syscall.Handle(db.lockfile.Fd()), 0, 1, 0, &syscall.Overlapped{})
|
||||
db.lockfile.Close()
|
||||
os.Remove(db.path+lockExt)
|
||||
return err
|
||||
}
|
||||
|
||||
// mmap memory maps a DB's data file.
|
||||
|
|
|
|||
File diff suppressed because it is too large
Load Diff
|
|
@ -1,8 +1,10 @@
|
|||
package bolt
|
||||
|
||||
import (
|
||||
"errors"
|
||||
"fmt"
|
||||
"hash/fnv"
|
||||
"log"
|
||||
"os"
|
||||
"runtime"
|
||||
"runtime/debug"
|
||||
|
|
@ -24,13 +26,14 @@ const magic uint32 = 0xED0CDAED
|
|||
// IgnoreNoSync specifies whether the NoSync field of a DB is ignored when
|
||||
// syncing changes to a file. This is required as some operating systems,
|
||||
// such as OpenBSD, do not have a unified buffer cache (UBC) and writes
|
||||
// must be synchronzied using the msync(2) syscall.
|
||||
// must be synchronized using the msync(2) syscall.
|
||||
const IgnoreNoSync = runtime.GOOS == "openbsd"
|
||||
|
||||
// Default values if not set in a DB instance.
|
||||
const (
|
||||
DefaultMaxBatchSize int = 1000
|
||||
DefaultMaxBatchDelay = 10 * time.Millisecond
|
||||
DefaultAllocSize = 16 * 1024 * 1024
|
||||
)
|
||||
|
||||
// DB represents a collection of buckets persisted to a file on disk.
|
||||
|
|
@ -83,11 +86,18 @@ type DB struct {
|
|||
// Do not change concurrently with calls to Batch.
|
||||
MaxBatchDelay time.Duration
|
||||
|
||||
// AllocSize is the amount of space allocated when the database
|
||||
// needs to create new pages. This is done to amortize the cost
|
||||
// of truncate() and fsync() when growing the data file.
|
||||
AllocSize int
|
||||
|
||||
path string
|
||||
file *os.File
|
||||
lockfile *os.File // windows only
|
||||
dataref []byte // mmap'ed readonly, write throws SEGV
|
||||
data *[maxMapSize]byte
|
||||
datasz int
|
||||
filesz int // current on disk file size
|
||||
meta0 *meta
|
||||
meta1 *meta
|
||||
pageSize int
|
||||
|
|
@ -145,6 +155,7 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) {
|
|||
// Set default values for later DB operations.
|
||||
db.MaxBatchSize = DefaultMaxBatchSize
|
||||
db.MaxBatchDelay = DefaultMaxBatchDelay
|
||||
db.AllocSize = DefaultAllocSize
|
||||
|
||||
flag := os.O_RDWR
|
||||
if options.ReadOnly {
|
||||
|
|
@ -167,7 +178,7 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) {
|
|||
// if !options.ReadOnly.
|
||||
// The database file is locked using the shared lock (more than one process may
|
||||
// hold a lock at the same time) otherwise (options.ReadOnly is set).
|
||||
if err := flock(db.file, !db.readOnly, options.Timeout); err != nil {
|
||||
if err := flock(db, mode, !db.readOnly, options.Timeout); err != nil {
|
||||
_ = db.close()
|
||||
return nil, err
|
||||
}
|
||||
|
|
@ -196,7 +207,7 @@ func Open(path string, mode os.FileMode, options *Options) (*DB, error) {
|
|||
}
|
||||
|
||||
// Memory map the data file.
|
||||
if err := db.mmap(0); err != nil {
|
||||
if err := db.mmap(options.InitialMmapSize); err != nil {
|
||||
_ = db.close()
|
||||
return nil, err
|
||||
}
|
||||
|
|
@ -271,7 +282,7 @@ func (db *DB) munmap() error {
|
|||
}
|
||||
|
||||
// mmapSize determines the appropriate size for the mmap given the current size
|
||||
// of the database. The minimum size is 1MB and doubles until it reaches 1GB.
|
||||
// of the database. The minimum size is 32KB and doubles until it reaches 1GB.
|
||||
// Returns an error if the new mmap size is greater than the max allowed.
|
||||
func (db *DB) mmapSize(size int) (int, error) {
|
||||
// Double the size from 32KB until 1GB.
|
||||
|
|
@ -369,6 +380,10 @@ func (db *DB) Close() error {
|
|||
}
|
||||
|
||||
func (db *DB) close() error {
|
||||
if !db.opened {
|
||||
return nil
|
||||
}
|
||||
|
||||
db.opened = false
|
||||
|
||||
db.freelist = nil
|
||||
|
|
@ -387,7 +402,9 @@ func (db *DB) close() error {
|
|||
// No need to unlock read-only file.
|
||||
if !db.readOnly {
|
||||
// Unlock the file.
|
||||
_ = funlock(db.file)
|
||||
if err := funlock(db); err != nil {
|
||||
log.Printf("bolt.Close(): funlock error: %s", err)
|
||||
}
|
||||
}
|
||||
|
||||
// Close the file descriptor.
|
||||
|
|
@ -406,11 +423,15 @@ func (db *DB) close() error {
|
|||
// will cause the calls to block and be serialized until the current write
|
||||
// transaction finishes.
|
||||
//
|
||||
// Transactions should not be depedent on one another. Opening a read
|
||||
// Transactions should not be dependent on one another. Opening a read
|
||||
// transaction and a write transaction in the same goroutine can cause the
|
||||
// writer to deadlock because the database periodically needs to re-mmap itself
|
||||
// as it grows and it cannot do that while a read transaction is open.
|
||||
//
|
||||
// If a long running read transaction (for example, a snapshot transaction) is
|
||||
// needed, you might want to set DB.InitialMmapSize to a large enough value
|
||||
// to avoid potential blocking of write transaction.
|
||||
//
|
||||
// IMPORTANT: You must close read-only transactions after you are finished or
|
||||
// else the database will not reclaim old pages.
|
||||
func (db *DB) Begin(writable bool) (*Tx, error) {
|
||||
|
|
@ -594,6 +615,136 @@ func (db *DB) View(fn func(*Tx) error) error {
|
|||
return nil
|
||||
}
|
||||
|
||||
// Batch calls fn as part of a batch. It behaves similar to Update,
|
||||
// except:
|
||||
//
|
||||
// 1. concurrent Batch calls can be combined into a single Bolt
|
||||
// transaction.
|
||||
//
|
||||
// 2. the function passed to Batch may be called multiple times,
|
||||
// regardless of whether it returns error or not.
|
||||
//
|
||||
// This means that Batch function side effects must be idempotent and
|
||||
// take permanent effect only after a successful return is seen in
|
||||
// caller.
|
||||
//
|
||||
// The maximum batch size and delay can be adjusted with DB.MaxBatchSize
|
||||
// and DB.MaxBatchDelay, respectively.
|
||||
//
|
||||
// Batch is only useful when there are multiple goroutines calling it.
|
||||
func (db *DB) Batch(fn func(*Tx) error) error {
|
||||
errCh := make(chan error, 1)
|
||||
|
||||
db.batchMu.Lock()
|
||||
if (db.batch == nil) || (db.batch != nil && len(db.batch.calls) >= db.MaxBatchSize) {
|
||||
// There is no existing batch, or the existing batch is full; start a new one.
|
||||
db.batch = &batch{
|
||||
db: db,
|
||||
}
|
||||
db.batch.timer = time.AfterFunc(db.MaxBatchDelay, db.batch.trigger)
|
||||
}
|
||||
db.batch.calls = append(db.batch.calls, call{fn: fn, err: errCh})
|
||||
if len(db.batch.calls) >= db.MaxBatchSize {
|
||||
// wake up batch, it's ready to run
|
||||
go db.batch.trigger()
|
||||
}
|
||||
db.batchMu.Unlock()
|
||||
|
||||
err := <-errCh
|
||||
if err == trySolo {
|
||||
err = db.Update(fn)
|
||||
}
|
||||
return err
|
||||
}
|
||||
|
||||
type call struct {
|
||||
fn func(*Tx) error
|
||||
err chan<- error
|
||||
}
|
||||
|
||||
type batch struct {
|
||||
db *DB
|
||||
timer *time.Timer
|
||||
start sync.Once
|
||||
calls []call
|
||||
}
|
||||
|
||||
// trigger runs the batch if it hasn't already been run.
|
||||
func (b *batch) trigger() {
|
||||
b.start.Do(b.run)
|
||||
}
|
||||
|
||||
// run performs the transactions in the batch and communicates results
|
||||
// back to DB.Batch.
|
||||
func (b *batch) run() {
|
||||
b.db.batchMu.Lock()
|
||||
b.timer.Stop()
|
||||
// Make sure no new work is added to this batch, but don't break
|
||||
// other batches.
|
||||
if b.db.batch == b {
|
||||
b.db.batch = nil
|
||||
}
|
||||
b.db.batchMu.Unlock()
|
||||
|
||||
retry:
|
||||
for len(b.calls) > 0 {
|
||||
var failIdx = -1
|
||||
err := b.db.Update(func(tx *Tx) error {
|
||||
for i, c := range b.calls {
|
||||
if err := safelyCall(c.fn, tx); err != nil {
|
||||
failIdx = i
|
||||
return err
|
||||
}
|
||||
}
|
||||
return nil
|
||||
})
|
||||
|
||||
if failIdx >= 0 {
|
||||
// take the failing transaction out of the batch. it's
|
||||
// safe to shorten b.calls here because db.batch no longer
|
||||
// points to us, and we hold the mutex anyway.
|
||||
c := b.calls[failIdx]
|
||||
b.calls[failIdx], b.calls = b.calls[len(b.calls)-1], b.calls[:len(b.calls)-1]
|
||||
// tell the submitter re-run it solo, continue with the rest of the batch
|
||||
c.err <- trySolo
|
||||
continue retry
|
||||
}
|
||||
|
||||
// pass success, or bolt internal errors, to all callers
|
||||
for _, c := range b.calls {
|
||||
if c.err != nil {
|
||||
c.err <- err
|
||||
}
|
||||
}
|
||||
break retry
|
||||
}
|
||||
}
|
||||
|
||||
// trySolo is a special sentinel error value used for signaling that a
|
||||
// transaction function should be re-run. It should never be seen by
|
||||
// callers.
|
||||
var trySolo = errors.New("batch function returned an error and should be re-run solo")
|
||||
|
||||
type panicked struct {
|
||||
reason interface{}
|
||||
}
|
||||
|
||||
func (p panicked) Error() string {
|
||||
if err, ok := p.reason.(error); ok {
|
||||
return err.Error()
|
||||
}
|
||||
return fmt.Sprintf("panic: %v", p.reason)
|
||||
}
|
||||
|
||||
func safelyCall(fn func(*Tx) error, tx *Tx) (err error) {
|
||||
defer func() {
|
||||
if p := recover(); p != nil {
|
||||
err = panicked{p}
|
||||
}
|
||||
}()
|
||||
return fn(tx)
|
||||
}
|
||||
|
||||
// Sync executes fdatasync() against the database file handle.
|
||||
//
|
||||
// This is not necessary under normal operation, however, if you use NoSync
|
||||
|
|
@ -660,6 +811,38 @@ func (db *DB) allocate(count int) (*page, error) {
|
|||
return p, nil
|
||||
}
|
||||
|
||||
// grow grows the size of the database to the given sz.
|
||||
func (db *DB) grow(sz int) error {
|
||||
// Ignore if the new size is less than available file size.
|
||||
if sz <= db.filesz {
|
||||
return nil
|
||||
}
|
||||
|
||||
// If the data is smaller than the alloc size then only allocate what's needed.
|
||||
// Once it goes over the allocation size then allocate in chunks.
|
||||
if db.datasz < db.AllocSize {
|
||||
sz = db.datasz
|
||||
} else {
|
||||
sz += db.AllocSize
|
||||
}
|
||||
|
||||
// Truncate and fsync to ensure file size metadata is flushed.
|
||||
// https://github.com/boltdb/bolt/issues/284
|
||||
if !db.NoGrowSync && !db.readOnly {
|
||||
if runtime.GOOS != "windows" {
|
||||
if err := db.file.Truncate(int64(sz)); err != nil {
|
||||
return fmt.Errorf("file resize error: %s", err)
|
||||
}
|
||||
}
|
||||
if err := db.file.Sync(); err != nil {
|
||||
return fmt.Errorf("file sync error: %s", err)
|
||||
}
|
||||
}
|
||||
|
||||
db.filesz = sz
|
||||
return nil
|
||||
}
|
||||
|
||||
func (db *DB) IsReadOnly() bool {
|
||||
return db.readOnly
|
||||
}
|
||||
|
|
@ -680,6 +863,16 @@ type Options struct {
|
|||
|
||||
// Sets the DB.MmapFlags flag before memory mapping the file.
|
||||
MmapFlags int
|
||||
|
||||
// InitialMmapSize is the initial mmap size of the database
|
||||
// in bytes. Read transactions won't block write transaction
|
||||
// if the InitialMmapSize is large enough to hold database mmap
|
||||
// size. (See DB.Begin for more information)
|
||||
//
|
||||
// If <=0, the initial map size is 0.
|
||||
// If initialMmapSize is smaller than the previous database size,
|
||||
// it takes no effect.
|
||||
InitialMmapSize int
|
||||
}
|
||||
|
||||
// DefaultOptions represent the options used if nil options are passed into Open().
|
||||
|
|
|
|||
|
|
@ -463,43 +463,6 @@ func (n *node) rebalance() {
|
|||
target = n.prevSibling()
|
||||
}
|
||||
|
||||
// If target node has extra nodes then just move one over.
|
||||
if target.numChildren() > target.minKeys() {
|
||||
if useNextSibling {
|
||||
// Reparent and move node.
|
||||
if child, ok := n.bucket.nodes[target.inodes[0].pgid]; ok {
|
||||
child.parent.removeChild(child)
|
||||
child.parent = n
|
||||
child.parent.children = append(child.parent.children, child)
|
||||
}
|
||||
n.inodes = append(n.inodes, target.inodes[0])
|
||||
target.inodes = target.inodes[1:]
|
||||
|
||||
// Update target key on parent.
|
||||
target.parent.put(target.key, target.inodes[0].key, nil, target.pgid, 0)
|
||||
target.key = target.inodes[0].key
|
||||
_assert(len(target.key) > 0, "rebalance(1): zero-length node key")
|
||||
} else {
|
||||
// Reparent and move node.
|
||||
if child, ok := n.bucket.nodes[target.inodes[len(target.inodes)-1].pgid]; ok {
|
||||
child.parent.removeChild(child)
|
||||
child.parent = n
|
||||
child.parent.children = append(child.parent.children, child)
|
||||
}
|
||||
n.inodes = append(n.inodes, inode{})
|
||||
copy(n.inodes[1:], n.inodes)
|
||||
n.inodes[0] = target.inodes[len(target.inodes)-1]
|
||||
target.inodes = target.inodes[:len(target.inodes)-1]
|
||||
}
|
||||
|
||||
// Update parent key for node.
|
||||
n.parent.put(n.key, n.inodes[0].key, nil, n.pgid, 0)
|
||||
n.key = n.inodes[0].key
|
||||
_assert(len(n.key) > 0, "rebalance(2): zero-length node key")
|
||||
|
||||
return
|
||||
}
|
||||
|
||||
// If both this node and the target node are too small then merge them.
|
||||
if useNextSibling {
|
||||
// Reparent all child nodes being moved.
|
||||
|
|
|
|||
|
|
@ -5,6 +5,7 @@ import (
|
|||
"io"
|
||||
"os"
|
||||
"sort"
|
||||
"strings"
|
||||
"time"
|
||||
"unsafe"
|
||||
)
|
||||
|
|
@ -168,6 +169,8 @@ func (tx *Tx) Commit() error {
|
|||
// Free the old root bucket.
|
||||
tx.meta.root.root = tx.root.root
|
||||
|
||||
opgid := tx.meta.pgid
|
||||
|
||||
// Free the freelist and allocate new pages for it. This will overestimate
|
||||
// the size of the freelist but not underestimate the size (which would be bad).
|
||||
tx.db.freelist.free(tx.meta.txid, tx.db.page(tx.meta.freelist))
|
||||
|
|
@ -182,6 +185,14 @@ func (tx *Tx) Commit() error {
|
|||
}
|
||||
tx.meta.freelist = p.id
|
||||
|
||||
// If the high water mark has moved up then attempt to grow the database.
|
||||
if tx.meta.pgid > opgid {
|
||||
if err := tx.db.grow(int(tx.meta.pgid+1) * tx.db.pageSize); err != nil {
|
||||
tx.rollback()
|
||||
return err
|
||||
}
|
||||
}
|
||||
|
||||
// Write dirty pages to disk.
|
||||
startTime = time.Now()
|
||||
if err := tx.write(); err != nil {
|
||||
|
|
@ -192,8 +203,17 @@ func (tx *Tx) Commit() error {
|
|||
// If strict mode is enabled then perform a consistency check.
|
||||
// Only the first consistency error is reported in the panic.
|
||||
if tx.db.StrictMode {
|
||||
if err, ok := <-tx.Check(); ok {
|
||||
panic("check fail: " + err.Error())
|
||||
ch := tx.Check()
|
||||
var errs []string
|
||||
for {
|
||||
err, ok := <-ch
|
||||
if !ok {
|
||||
break
|
||||
}
|
||||
errs = append(errs, err.Error())
|
||||
}
|
||||
if len(errs) > 0 {
|
||||
panic("check fail: " + strings.Join(errs, "\n"))
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -285,14 +305,36 @@ func (tx *Tx) WriteTo(w io.Writer) (n int64, err error) {
|
|||
if err != nil {
|
||||
return 0, err
|
||||
}
|
||||
defer f.Close()
|
||||
defer func() { _ = f.Close() }()
|
||||
|
||||
// Copy the meta pages.
|
||||
tx.db.metalock.Lock()
|
||||
n, err = io.CopyN(w, f, int64(tx.db.pageSize*2))
|
||||
tx.db.metalock.Unlock()
|
||||
// Generate a meta page. We use the same page data for both meta pages.
|
||||
buf := make([]byte, tx.db.pageSize)
|
||||
page := (*page)(unsafe.Pointer(&buf[0]))
|
||||
page.flags = metaPageFlag
|
||||
*page.meta() = *tx.meta
|
||||
|
||||
// Write meta 0.
|
||||
page.id = 0
|
||||
page.meta().checksum = page.meta().sum64()
|
||||
nn, err := w.Write(buf)
|
||||
n += int64(nn)
|
||||
if err != nil {
|
||||
return n, fmt.Errorf("meta copy: %s", err)
|
||||
return n, fmt.Errorf("meta 0 copy: %s", err)
|
||||
}
|
||||
|
||||
// Write meta 1 with a lower transaction id.
|
||||
page.id = 1
|
||||
page.meta().txid -= 1
|
||||
page.meta().checksum = page.meta().sum64()
|
||||
nn, err = w.Write(buf)
|
||||
n += int64(nn)
|
||||
if err != nil {
|
||||
return n, fmt.Errorf("meta 1 copy: %s", err)
|
||||
}
|
||||
|
||||
// Move past the meta pages in the file.
|
||||
if _, err := f.Seek(int64(tx.db.pageSize*2), os.SEEK_SET); err != nil {
|
||||
return n, fmt.Errorf("seek: %s", err)
|
||||
}
|
||||
|
||||
// Copy data pages.
|
||||
|
|
@ -505,7 +547,7 @@ func (tx *Tx) writeMeta() error {
|
|||
}
|
||||
|
||||
// page returns a reference to the page with a given id.
|
||||
// If page has been written to then a temporary bufferred page is returned.
|
||||
// If page has been written to then a temporary buffered page is returned.
|
||||
func (tx *Tx) page(id pgid) *page {
|
||||
// Check the dirty pages first.
|
||||
if tx.pages != nil {
|
||||
|
|
|
|||
Loading…
Reference in New Issue