Add ARC caching and bloom filter for blockstorage #2885
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,175 @@ | ||
| package blockstore | ||
|
|
||
| import ( | ||
| "github.com/ipfs/go-ipfs/blocks" | ||
| key "github.com/ipfs/go-ipfs/blocks/key" | ||
| lru "gx/ipfs/QmVYxfoJQiZijTgPNHCHgHELvQpbsJNTg6Crmc3dQkj3yy/golang-lru" | ||
| bloom "gx/ipfs/QmWQ2SJisXwcCLsUXLwYCKSfyExXjFRW2WbBH5sqCUnwX5/bbloom" | ||
| context "gx/ipfs/QmZy2y8t9zQH2a1b8q2ZSLKp17ATuJoCNxxyMFG5qFExpt/go-net/context" | ||
| ds "gx/ipfs/QmfQzVugPq1w5shWRcLWSeiHF4a2meBX7yVD8Vw7GWJM9o/go-datastore" | ||
| ) | ||
|
|
||
| // BloomCached returns Blockstore that caches Has requests using Bloom filter | ||
| // Size is size of bloom filter in bytes | ||
| func BloomCached(bs Blockstore, bloomSize, lruSize int) (*bloomcache, error) { | ||
| bl, err := bloom.New(float64(bloomSize), float64(7)) | ||
| if err != nil { | ||
| return nil, err | ||
| } | ||
| arc, err := lru.NewARC(lruSize) | ||
| if err != nil { | ||
| return nil, err | ||
| } | ||
| bc := &bloomcache{blockstore: bs, bloom: bl, arc: arc} | ||
| bc.Invalidate() | ||
| go bc.Rebuild() | ||
|
|
||
| return bc, nil | ||
| } | ||
|
|
||
| type bloomcache struct { | ||
| bloom *bloom.Bloom | ||
| active bool | ||
|
|
||
| arc *lru.ARCCache | ||
| // This chan is only used for testing to wait for bloom to enable | ||
| rebuildChan chan struct{} | ||
| blockstore Blockstore | ||
|
|
||
| // Statistics | ||
| hits uint64 | ||
| misses uint64 | ||
| } | ||
|
|
||
| func (b *bloomcache) Invalidate() { | ||
| b.rebuildChan = make(chan struct{}) | ||
| b.active = false | ||
| } | ||
|
|
||
| func (b *bloomcache) BloomActive() bool { | ||
| return b.active | ||
| } | ||
|
|
||
| func (b *bloomcache) Rebuild() { | ||
| ctx := context.TODO() | ||
| evt := log.EventBegin(ctx, "bloomcache.Rebuild") | ||
| defer evt.Done() | ||
|
|
||
| ch, err := b.blockstore.AllKeysChan(ctx) | ||
| if err != nil { | ||
| log.Errorf("AllKeysChan failed in bloomcache rebuild with: %v", err) | ||
| return | ||
| } | ||
| for key := range ch { | ||
| b.bloom.AddTS([]byte(key)) // Use binary key, the more compact the better | ||
| } | ||
| close(b.rebuildChan) | ||
| b.active = true | ||
| } | ||
|
|
||
| func (b *bloomcache) DeleteBlock(k key.Key) error { | ||
| if has, ok := b.hasCached(k); ok && !has { | ||
| return ErrNotFound | ||
| } | ||
|
|
||
| b.arc.Remove(k) // Invalidate cache before deleting. | ||
| err := b.blockstore.DeleteBlock(k) | ||
| if err == nil { | ||
|
There was a problem hiding this comment. switch err {
case nil:
//
case ds.ErrNotFound, ErrNotFound:
//
default:
return err
} |
||
| b.arc.Add(k, false) | ||
| } else if err == ds.ErrNotFound || err == ErrNotFound { | ||
| b.arc.Add(k, false) | ||
| return ErrNotFound | ||
| } | ||
| return err | ||
| } | ||
|
|
||
| // if ok == false has is inconclusive | ||
| // if ok == true then has respons to question: is it contained | ||
| func (b *bloomcache) hasCached(k key.Key) (has bool, ok bool) { | ||
| if k == "" { | ||
|
There was a problem hiding this comment. what? we always have the empty key? There was a problem hiding this comment. It was there in the write_cache, I just mimicked it. There was a problem hiding this comment. oh weird... i wonder what the reasoning there was |
||
| return true, true | ||
| } | ||
| if b.active { | ||
|
There was a problem hiding this comment. this field needs to be locked around There was a problem hiding this comment. Does it? I would prefer not to introduce locking here as it is quite slow when compared with speed of bloom. It would only introduce race if bloom was being deactivated (it isn't in yet, but that doesn't matter) and the pointer to I will leave a comment about that if it seems good to you. There was a problem hiding this comment. rebuild never gets called outside of the constructor currently (which i'm not super keen on), but even the one call in the constructor will cause a race. What if methods are being called right as its being set to true? setting a boolean isnt guaranteed to be atomic There was a problem hiding this comment. Yes, but when it is set from false to true then race is that |
||
| blr := b.bloom.HasTS([]byte(k)) | ||
| if blr == false { // not contained in bloom is only conclusive answer bloom gives | ||
| return blr, true | ||
|
There was a problem hiding this comment. lets hardcode |
||
| } | ||
| } | ||
| h, ok := b.arc.Get(k) | ||
| if ok { | ||
| return h.(bool), ok | ||
| } else { | ||
| return false, ok | ||
| } | ||
| } | ||
|
|
||
| func (b *bloomcache) Has(k key.Key) (bool, error) { | ||
| if has, ok := b.hasCached(k); ok { | ||
| return has, nil | ||
| } | ||
|
|
||
| res, err := b.blockstore.Has(k) | ||
| if err == nil { | ||
| b.arc.Add(k, res) | ||
| } | ||
| return res, err | ||
| } | ||
|
|
||
| func (b *bloomcache) Get(k key.Key) (blocks.Block, error) { | ||
| if has, ok := b.hasCached(k); ok && !has { | ||
| return nil, ErrNotFound | ||
| } | ||
|
|
||
| bl, err := b.blockstore.Get(k) | ||
| if bl == nil && err == ErrNotFound { | ||
| b.arc.Add(k, false) | ||
| } else if bl != nil { | ||
| b.arc.Add(k, true) | ||
| } | ||
| return bl, err | ||
| } | ||
|
|
||
| func (b *bloomcache) Put(bl blocks.Block) error { | ||
| if has, ok := b.hasCached(bl.Key()); ok && has { | ||
| return nil | ||
| } | ||
|
|
||
| err := b.blockstore.Put(bl) | ||
| if err == nil { | ||
| b.bloom.AddTS([]byte(bl.Key())) | ||
| b.arc.Add(bl.Key(), true) | ||
| } | ||
| return err | ||
| } | ||
|
|
||
| func (b *bloomcache) PutMany(bs []blocks.Block) error { | ||
| var good []blocks.Block | ||
| for _, block := range bs { | ||
| if has, ok := b.hasCached(block.Key()); !ok || (ok && !has) { | ||
| good = append(good, block) | ||
| } | ||
| } | ||
| err := b.blockstore.PutMany(bs) | ||
| if err == nil { | ||
| for _, block := range bs { | ||
| b.bloom.AddTS([]byte(block.Key())) | ||
| } | ||
| } | ||
| return err | ||
| } | ||
|
|
||
| func (b *bloomcache) AllKeysChan(ctx context.Context) (<-chan key.Key, error) { | ||
| return b.blockstore.AllKeysChan(ctx) | ||
| } | ||
|
|
||
| func (b *bloomcache) GCLock() Unlocker { | ||
| return b.blockstore.(GCBlockstore).GCLock() | ||
| } | ||
|
|
||
| func (b *bloomcache) PinLock() Unlocker { | ||
| return b.blockstore.(GCBlockstore).PinLock() | ||
| } | ||
|
|
||
| func (b *bloomcache) GCRequested() bool { | ||
| return b.blockstore.(GCBlockstore).GCRequested() | ||
| } | ||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
we should try and get a context passed in, maybe from the constructor?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Seems like a good idea.