Documentation
¶
Index ¶
- type Hash
- type KVRead
- type KVWrite
- type MerkleSummary
- type MerkleTreeLevel
- type NsReadWriteSet
- type RWSet
- func (rws *RWSet) AddToRangeQuerySet(ns string, rqi *RangeQueryInfo)
- func (rws *RWSet) AddToReadSet(ns string, key string, version *version.Height)
- func (rws *RWSet) AddToWriteSet(ns string, key string, value []byte)
- func (rws *RWSet) GetFromWriteSet(ns string, key string) ([]byte, bool)
- func (rws *RWSet) GetTxReadWriteSet() *TxReadWriteSet
- type RangeQueryInfo
- type RangeQueryResultsHelper
- type TxReadWriteSet
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type KVRead ¶
KVRead - a tuple of key and its version at the time of transaction simulation
type KVWrite ¶
KVWrite - a tuple of key and it's value that a transaction wants to set during simulation. In addition, IsDelete is set to true iff the operation performed on the key is a delete operation
func NewKVWrite ¶
NewKVWrite constructs a new `KVWrite`
type MerkleSummary ¶
type MerkleSummary struct {
MaxDegree int
MaxLevel MerkleTreeLevel
MaxLevelHashes []Hash
}
MerkleSummary encloses the summary of the merkle tree that consists of the hashes of the results of a range query. This allows to reduce the size of RWSet in the presence of range query results by storing certain hashes instead of actual results.
func (*MerkleSummary) Equal ¶
func (ms *MerkleSummary) Equal(anotherMS *MerkleSummary) bool
Equal verifies whether the give MerkleSummary is equals to this
type MerkleTreeLevel ¶
type MerkleTreeLevel int
MerkleTreeLevel used for representing a level of the merkle tree
type NsReadWriteSet ¶
type NsReadWriteSet struct {
NameSpace string
Reads []*KVRead
Writes []*KVWrite
RangeQueriesInfo []*RangeQueryInfo
}
NsReadWriteSet - a collection of all the reads and writes that belong to a common namespace
func (*NsReadWriteSet) Marshal ¶
func (nsRW *NsReadWriteSet) Marshal(buf *proto.Buffer) error
Marshal serializes a `NsReadWriteSet`
func (*NsReadWriteSet) String ¶
func (nsRW *NsReadWriteSet) String() string
String prints a `NsReadWriteSet`
type RWSet ¶
type RWSet struct {
// contains filtered or unexported fields
}
RWSet maintains the read-write set
func (*RWSet) AddToRangeQuerySet ¶
func (rws *RWSet) AddToRangeQuerySet(ns string, rqi *RangeQueryInfo)
AddToRangeQuerySet adds a range query info for performing phantom read validation
func (*RWSet) AddToReadSet ¶
AddToReadSet adds a key and corresponding version to the read-set
func (*RWSet) AddToWriteSet ¶
AddToWriteSet adds a key and value to the write-set
func (*RWSet) GetFromWriteSet ¶
GetFromWriteSet return the value of a key from the write-set
func (*RWSet) GetTxReadWriteSet ¶
func (rws *RWSet) GetTxReadWriteSet() *TxReadWriteSet
GetTxReadWriteSet returns the read-write set in the form that can be serialized
type RangeQueryInfo ¶
type RangeQueryInfo struct {
StartKey string
EndKey string
ItrExhausted bool
Results []*KVRead
ResultHash *MerkleSummary
}
RangeQueryInfo captures a range query executed by a transaction and the tuples <key,version> that are read by the transaction This it to be used to perform a phantom-read validation during commit
func (*RangeQueryInfo) Marshal ¶
func (rqi *RangeQueryInfo) Marshal(buf *proto.Buffer) error
Marshal serializes a `RangeQueryInfo`
func (*RangeQueryInfo) String ¶
func (rqi *RangeQueryInfo) String() string
String prints a range query info
type RangeQueryResultsHelper ¶
type RangeQueryResultsHelper struct {
// contains filtered or unexported fields
}
RangeQueryResultsHelper helps preparing range query results for phantom items detection during validation. The results are expected to be fed as they are being iterated over. If the `hashingEnabled` is set to true, a merkle tree is built of the hashes over the results. The merkle tree helps reducing the size of the RWSet which otherwise would need to store all the raw KVReads
The mental model of the tree can be described as below: All the results are treated as leaf nodes (level 0) of the tree. Next up level of the tree is built by collecting 'maxDegree + 1' items from the previous level and hashing the entire collection. Further upper levels of the tree are built in similar manner however the only difference is that unlike level-0 (where collection consists of raw KVReads), collection at level 1 and above, consists of the hashes (of the collection of previous level). This is repeated until we reach at a level where we are left with the number of items less than or equals to `maxDegree`. In the last collection, the number of items can be less than 'maxDegree' (except if this is the only collection at the given level).
As a result, if the number of total input results are less than or equals to 'maxDegree', no hashing is performed at all. And the final output of the computation is either the collection of raw results (if less that or equals to 'maxDegree') or a collection of hashes (that or equals to 'maxDegree') at some level in the tree.
`AddResult` function should be invoke to supply the next result and at the end `Done` function should be invoked. The `Done` function does the final processing and returns the final output
func NewRangeQueryResultsHelper ¶
func NewRangeQueryResultsHelper(enableHashing bool, maxDegree int) (*RangeQueryResultsHelper, error)
NewRangeQueryResultsHelper constructs a RangeQueryResultsHelper
func (*RangeQueryResultsHelper) AddResult ¶
func (helper *RangeQueryResultsHelper) AddResult(kvRead *KVRead) error
AddResult adds a new query result for processing. Put the result into the list of pending results. If the number of pending results exceeds `maxDegree`, consume the results for incrementally update the merkle tree
func (*RangeQueryResultsHelper) Done ¶
func (helper *RangeQueryResultsHelper) Done() ([]*KVRead, *MerkleSummary, error)
Done processes any pending results if needed This returns the final pending results (i.e., []*KVRead) and hashes of the results (i.e., *MerkleSummary) Only one of these two will be non-nil (except when no results are ever added). `MerkleSummary` will be nil if and only if either `enableHashing` is set to false or the number of total results are less than `maxDegree`
func (*RangeQueryResultsHelper) GetMerkleSummary ¶
func (helper *RangeQueryResultsHelper) GetMerkleSummary() *MerkleSummary
GetMerkleSummary return the current state of the MerkleSummary This intermediate state of the merkle tree helps during validation to detect a mismatch early on. That helps by not requiring to build the complete merkle tree during validation if there is a mismatch in early portion of the result-set.
type TxReadWriteSet ¶
type TxReadWriteSet struct {
NsRWs []*NsReadWriteSet
}
TxReadWriteSet - a collection of all the reads and writes collected as a result of a transaction simulation
func (*TxReadWriteSet) Marshal ¶
func (txRW *TxReadWriteSet) Marshal() ([]byte, error)
Marshal serializes a `TxReadWriteSet`
func (*TxReadWriteSet) String ¶
func (txRW *TxReadWriteSet) String() string
String prints a `TxReadWriteSet`
func (*TxReadWriteSet) Unmarshal ¶
func (txRW *TxReadWriteSet) Unmarshal(b []byte) error
Unmarshal deserializes a `TxReadWriteSet`
Source Files