README
¶
Streaming
Streaming is a new protocol of the swarm bzz bundle of protocols. This protocol provides the basic logic for chunk-based data flow. It implements simple retrieve requests and delivery using priority queue. A data exchange stream is a directional flow of chunks between peers. The source of datachunks is the upstream, the receiver is called the downstream peer. Each streaming protocol defines an outgoing streamer and an incoming streamer, the former installing on the upstream, the latter on the downstream peer.
Subscribe on StreamerPeer launches an incoming streamer that sends a subscribe msg upstream. The streamer on the upstream peer handles the subscribe msg by installing the relevant outgoing streamer . The modules now engage in a process of upstream sending a sequence of hashes of chunks downstream (OfferedHashesMsg). The downstream peer evaluates which hashes are needed and get it delivered by sending back a msg (WantedHashesMsg).
Historical syncing is supported - currently not the right abstraction -- state kept across sessions by saving a series of intervals after their last batch actually arrived.
Live streaming is also supported, by starting session from the first item after the subscription.
Provable data exchange. In case a stream represents a swarm document's data layer or higher level chunks, streaming up to a certain index is always provable. It saves on sending intermediate chunks.
Using the streamer logic, various stream types are easy to implement:
- light node requests:
- url lookup with offset
- document download
- document upload
- syncing
- live session syncing
- historical syncing
- simple retrieve requests and deliveries
- swarm feeds streams
- receipting for finger pointing
Syncing
Syncing is the process that makes sure storer nodes end up storing all and only the chunks that are requested from them.
Requirements
- eventual consistency: so each chunk historical should be syncable
- since the same chunk can and will arrive from many peers, (network traffic should be optimised, only one transfer of data per chunk)
- explicit request deliveries should be prioritised higher than recent chunks received during the ongoing session which in turn should be higher than historical chunks.
- insured chunks should get receipted for finger pointing litigation, the receipts storage should be organised efficiently, upstream peer should also be able to find these receipts for a deleted chunk easily to refute their challenge.
- syncing should be resilient to cut connections, metadata should be persisted that keep track of syncing state across sessions, historical syncing state should survive restart
- extra data structures to support syncing should be kept at minimum
- syncing is not organized separately for chunk types (Swarm feed updates v regular content chunk)
- various types of streams should have common logic abstracted
Syncing is now entirely mediated by the localstore, ie., no processes or memory leaks due to network contention. When a new chunk is stored, its chunk hash is index by proximity bin
peers syncronise by getting the chunks closer to the downstream peer than to the upstream one. Consequently peers just sync all stored items for the kad bin the receiving peer falls into. The special case of nearest neighbour sets is handled by the downstream peer indicating they want to sync all kademlia bins with proximity equal to or higher than their depth.
This sync state represents the initial state of a sync connection session. Retrieval is dictated by downstream peers simply using a special streamer protocol.
Syncing chunks created during the session by the upstream peer is called live session syncing while syncing of earlier chunks is historical syncing.
Once the relevant chunk is retrieved, downstream peer looks up all hash segments in its localstore
and sends to the upstream peer a message with a a bitvector to indicate
missing chunks (e.g., for chunk k, hash with chunk internal index which case )
new items. In turn upstream peer sends the relevant chunk data alongside their index.
On sending chunks there is a priority queue system. If during looking up hashes in its localstore, downstream peer hits on an open request then a retrieve request is sent immediately to the upstream peer indicating that no extra round of checks is needed. If another peers syncer hits the same open request, it is slightly unsafe to not ask that peer too: if the first one disconnects before delivering or fails to deliver and therefore gets disconnected, we should still be able to continue with the other. The minimum redundant traffic coming from such simultaneous eventualities should be sufficiently rare not to warrant more complex treatment.
Session syncing involves downstream peer to request a new state on a bin from upstream. using the new state, the range (of chunks) between the previous state and the new one are retrieved and chunks are requested identical to the historical case. After receiving all the missing chunks from the new hashes, downstream peer will request a new range. If this happens before upstream peer updates a new state, we say that session syncing is live or the two peers are in sync. In general the time interval passed since downstream peer request up to the current session cursor is a good indication of a permanent (probably increasing) lag.
If there is no historical backlog, and downstream peer has an acceptable 'last synced' tag, then it is said to be fully synced with the upstream peer. If a peer is fully synced with all its storer peers, it can advertise itself as globally fully synced.
The downstream peer persists the record of the last synced offset. When the two peers disconnect and reconnect syncing can start from there. This situation however can also happen while historical syncing is not yet complete. Effectively this means that the peer needs to persist a record of an arbitrary array of offset ranges covered.
Delivery requests
once the appropriate ranges of the hashstream are retrieved and buffered, downstream peer just scans the hashes, looks them up in localstore, if not found, create a request entry. The range is referenced by the chunk index. Alongside the name (indicating the stream, e.g., content chunks for bin 6) and the range downstream peer sends a 128 long bitvector indicating which chunks are needed. Newly created requests are satisfied bound together in a waitgroup which when done, will promptt sending the next one. to be able to do check and storage concurrently, we keep a buffer of one, we start with two batches of hashes. If there is nothing to give, upstream peers SetNextBatch is blocking. Subscription ends with an unsubscribe. which removes the syncer from the map.
Canceling requests (for instance the late chunks of an erasure batch) should be a chan closed on the request
Simple request is also a subscribe different streaming protocols are different p2p protocols with same message types. the constructor is the Run function itself. which takes a streamerpeer as argument
provable streams
The swarm hash over the hash stream has many advantages. It implements a provable data transfer and provide efficient storage for receipts in the form of inclusion proofs useable for finger pointing litigation. When challenged on a missing chunk, upstream peer will provide an inclusion proof of a chunk hash against the state of the sync stream. In order to be able to generate such an inclusion proof, upstream peer needs to store the hash index (counting consecutive hash-size segments) alongside the chunk data and preserve it even when the chunk data is deleted until the chunk is no longer insured. if there is no valid insurance on the files the entry may be deleted. As long as the chunk is preserved, no takeover proof will be needed since the node can respond to any challenge. However, once the node needs to delete an insured chunk for capacity reasons, a receipt should be available to refute the challenge by finger pointing to a downstream peer. As part of the deletion protocol then, hashes of insured chunks to be removed are pushed to an infinite stream for every bin.
Downstream peer on the other hand needs to make sure that they can only be finger pointed about a chunk they did receive and store. For this the check of a state should be exhaustive. If historical syncing finishes on one state, all hashes before are covered, no surprises. In other words historical syncing this process is self verifying. With session syncing however, it is not enough to check going back covering the range from old offset to new. Continuity (i.e., that the new state is extension of the old) needs to be verified: after downstream peer reads the range into a buffer, it appends the buffer the last known state at the last known offset and verifies the resulting hash matches the latest state. Past intervals of historical syncing are checked via the session root. Upstream peer signs the states, downstream peers can use as handover proofs. Downstream peers sign off on a state together with an initial offset.
Once historical syncing is complete and the session does not lag, downstream peer only preserves the latest upstream state and store the signed version.
Upstream peer needs to keep the latest takeover states: each deleted chunk's hash should be covered by takeover proof of at least one peer. If historical syncing is complete, upstream peer typically will store only the latest takeover proof from downstream peer. Crucially, the structure is totally independent of the number of peers in the bin, so it scales extremely well.
implementation
The simplest protocol just involves upstream peer to prefix the key with the kademlia proximity order (say 0-15 or 0-31) and simply iterate on index per bin when syncing with a peer.
priority queues are used for sending chunks so that user triggered requests should be responded to first, session syncing second, and historical with lower priority. The request on chunks remains implemented as a dataless entry in the memory store. The lifecycle of this object should be more carefully thought through, ie., when it fails to retrieve it should be removed.
Documentation
¶
Index ¶
- Constants
- Variables
- func Label(e *entry) string
- func LogAddrs(nns [][]byte) string
- func NewPeerPotMap(neighbourhoodSize int, addrs [][]byte) map[string]*PeerPot
- func NotifyDepth(depth uint8, kad *Kademlia)
- func NotifyPeer(p *BzzAddr, k *Kademlia)
- type Bzz
- func (b *Bzz) APIs() []rpc.API
- func (b *Bzz) GetOrCreateHandshake(peerID enode.ID) (*HandshakeMsg, bool)
- func (b *Bzz) NodeInfo() interface{}
- func (b *Bzz) Protocols() []p2p.Protocol
- func (b *Bzz) RunProtocol(spec *protocols.Spec, run func(*BzzPeer) error) func(*p2p.Peer, p2p.MsgReadWriter) error
- func (b *Bzz) UpdateLocalAddr(byteaddr []byte) *BzzAddr
- type BzzAddr
- type BzzConfig
- type BzzPeer
- type Fetcher
- type FetcherFactory
- type HandshakeMsg
- type Health
- type Hive
- type HiveParams
- type KadParams
- type Kademlia
- func (k *Kademlia) AddrCountC() <-chan int
- func (k *Kademlia) BaseAddr() []byte
- func (k *Kademlia) EachAddr(base []byte, o int, f func(*BzzAddr, int) bool)
- func (k *Kademlia) EachConn(base []byte, o int, f func(*Peer, int) bool)
- func (k *Kademlia) Healthy(pp *PeerPot) *Health
- func (k *Kademlia) NeighbourhoodDepth() (depth int)
- func (k *Kademlia) NeighbourhoodDepthC() <-chan int
- func (k *Kademlia) Off(p *Peer)
- func (k *Kademlia) On(p *Peer) (uint8, bool)
- func (k *Kademlia) Register(peers ...*BzzAddr) error
- func (k *Kademlia) String() string
- func (k *Kademlia) SuggestPeer() (suggestedPeer *BzzAddr, saturationDepth int, changed bool)
- type Peer
- type PeerPot
- type Request
- type RequestFunc
Constants ¶
const (
DefaultNetworkID = 3
)
Variables ¶
var BzzSpec = &protocols.Spec{ Name: "bzz", Version: 8, MaxMsgSize: 10 * 1024 * 1024, Messages: []interface{}{ HandshakeMsg{}, }, }
bzzspec是通用群握手的规范
var DiscoverySpec = &protocols.Spec{
Name: "hive",
Version: 8,
MaxMsgSize: 10 * 1024 * 1024,
Messages: []interface{}{
peersMsg{},
subPeersMsg{},
},
}
discovery spec是bzz discovery子协议的规范
var Pof = pot.DefaultPof(256)
var RequestTimeout = 10 * time.Second
考虑跳过对等机的时间。 也用于流传送。
Functions ¶
func NewPeerPotMap ¶
newpeerpotmap用键创建一个pot记录的映射*bzzaddr 作为地址的十六进制表示。 使用通过的卡德米利亚的邻里大小 仅用于测试 TODO移动到单独的测试工具文件
Types ¶
type Bzz ¶
bzz是swarm协议包
func NewBzz ¶
func NewBzz(config *BzzConfig, kad *Kademlia, store state.Store, streamerSpec *protocols.Spec, streamerRun func(*BzzPeer) error) *Bzz
Newzz是Swarm协议的构造者 争论 *BZZ配置 *覆盖驱动程序 *对等存储
func (*Bzz) GetOrCreateHandshake ¶
func (b *Bzz) GetOrCreateHandshake(peerID enode.ID) (*HandshakeMsg, bool)
gethandshake返回peerid远程对等机发送的bzz handshake
func (*Bzz) RunProtocol ¶
func (b *Bzz) RunProtocol(spec *protocols.Spec, run func(*BzzPeer) error) func(*p2p.Peer, p2p.MsgReadWriter) error
runprotocol是swarm子协议的包装器 返回可分配给p2p.protocol run字段的p2p协议运行函数。 争论： *P2P协议规范 *以bzzpeer为参数运行函数 此运行函数用于在协议会话期间阻塞 返回时，会话终止，对等端断开连接。 协议等待BZZ握手被协商 bzzpeer上的覆盖地址是通过远程握手设置的。
func (*Bzz) UpdateLocalAddr ¶
updateLocalAddr更新正在运行的节点的参考底图地址
type BzzConfig ¶
type BzzConfig struct {
OverlayAddr []byte //覆盖网络的基址
UnderlayAddr []byte //节点的参考底图地址
HiveParams *HiveParams
NetworkID uint64
LightNode bool
}
bzzconfig捕获配置单元使用的配置参数
type BzzPeer ¶
type BzzPeer struct {
*protocols.Peer //表示联机对等机的连接
*BzzAddr //远程地址->实现addr interface=protocols.peer
LightNode bool
// contains filtered or unexported fields
}
bzz peer是协议的bzz协议视图。peer（本身是p2p.peer的扩展） 实现对等接口和所有接口对等实现：addr、overlaypeer
func NewBzzPeer ¶
type Fetcher ¶
type Fetcher struct {
// contains filtered or unexported fields
}
当在本地找不到块时，将创建提取程序。它启动一次请求处理程序循环，然后 在所有活动请求完成之前保持活动状态。这可能发生： 1。或者因为块被传递 2。或者因为请求者取消/超时 获取器在完成后自行销毁。 TODO:取消终止后的所有转发请求
func NewFetcher ¶
func NewFetcher(addr storage.Address, rf RequestFunc, skipCheck bool) *Fetcher
new fetcher使用给定的请求函数为给定的块地址创建一个新的fetcher。
type FetcherFactory ¶
type FetcherFactory struct {
// contains filtered or unexported fields
}
FetcherFactory是用请求函数初始化的，可以创建Fetcher
func NewFetcherFactory ¶
func NewFetcherFactory(request RequestFunc, skipCheck bool) *FetcherFactory
NewFetcherFactory接受请求函数并跳过检查参数并创建FetcherFactory
type HandshakeMsg ¶
type Health ¶
type Health struct {
KnowNN bool //节点是否知道其所有邻居
CountKnowNN int //已知邻居数量
MissingKnowNN [][]byte //我们应该知道哪些邻居，但我们不知道
ConnectNN bool //节点是否连接到其所有邻居
CountConnectNN int //连接到的邻居数量
MissingConnectNN [][]byte //我们应该和哪个邻居有联系，但我们没有
Saturated bool //我们是否与所有我们想联系的同龄人建立了联系
Hive string
}
卡德米利亚的健康状况 仅用于测试
type Hive ¶
type Hive struct {
*HiveParams //设置
*Kademlia //覆盖连接驱动程序
Store state.Store //存储接口，用于跨会话保存对等端
// contains filtered or unexported fields
}
蜂巢管理群节点的网络连接
func NewHive ¶
func NewHive(params *HiveParams, kad *Kademlia, store state.Store) *Hive
new hive构造新的hive hiveparams:配置参数 Kademlia：使用网络拓扑的连接驱动程序 Statestore：保存会话之间的对等点
func (*Hive) NodeInfo ¶
func (h *Hive) NodeInfo() interface{}
p2p.server rpc接口使用nodeinfo函数显示 协议特定的节点信息
type HiveParams ¶
type HiveParams struct {
Discovery bool //如果不想发现
PeersBroadcastSetSize uint8 //中继时要使用多少对等点
MaxPeersPerRequest uint8 //对等地址批的最大大小
KeepAliveInterval time.Duration
}
hiveparams保存配置选项以进行配置
type KadParams ¶
type KadParams struct {
// 可调参数
MaxProxDisplay int //表显示的行数
NeighbourhoodSize int //最近邻核最小基数
MinBinSize int //一行中的最小对等数
MaxBinSize int //修剪前一行中的最大对等数
RetryInterval int64 //对等机首次重新拨号前的初始间隔
RetryExponent int //用指数乘以重试间隔
MaxRetries int //重拨尝试的最大次数
//制裁或阻止建议同伴的职能
Reachable func(*BzzAddr) bool `json:"-"`
}
kadparams保存kademlia的配置参数
type Kademlia ¶
type Kademlia struct {
*KadParams //Kademlia配置参数
// contains filtered or unexported fields
}
Kademlia是一个活动对等端表和一个已知对等端数据库（节点记录）
func NewKademlia ¶
newkademlia为基地址addr创建一个kademlia表 参数与参数相同 如果params为nil，则使用默认值
func (*Kademlia) AddrCountC ¶
addrCountc返回发送新的 每次更改的地址计数值。 不从返回通道接收将阻止寄存器功能 地址计数值更改时。
func (*Kademlia) EachAddr ¶
用（base，po，f）调用的eachaddr是一个迭代器，将f应用于每个已知的对等端 从底部测量，接近顺序为O或更低 如果基为零，则使用Kademlia基地址
func (*Kademlia) EachConn ¶
eachconn是一个带有args（base、po、f）的迭代器，将f应用于每个活动对等端 从基地测量，接近订单为po或更低 如果基为零，则使用Kademlia基地址
func (*Kademlia) Healthy ¶
健康报告Kademlia连接性的健康状态
peerpot参数提供了网络的全知视图 结果健康对象是 问题中的卡德米利亚（接受者）的实际组成是什么，以及 当我们考虑到我们对网络的所有了解时，应该是什么情况呢？
仅用于测试
func (*Kademlia) NeighbourhoodDepth ¶
Neighbourhooddepth返回壶的深度，请参见壶的深度
func (*Kademlia) NeighbourhoodDepthC ¶
Neighbourhooddepthc返回发送新Kademlia的频道 每一次变化的邻里深度。 不从返回通道接收将阻塞功能 当邻近深度改变时。 托多：为什么要导出它，如果应该的话；为什么我们不能有多个订户？
type Peer ¶
type Peer struct {
*BzzPeer
// contains filtered or unexported fields
}
Peer包装BZZPeer并嵌入Kademlia覆盖连接驱动程序
func (*Peer) NotifyDepth ¶
notifydepth向接收者发送一个子程序msg，通知他们 饱和深度的变化
func (*Peer) NotifyPeer ¶
如果出现以下情况，notifypeer将通知远程节点（收件人）有关对等机的信息： 对等方的采购订单在收件人的广告深度内 或者对方比自己更接近对方 除非在连接会话期间已通知
Source Files
Directories