Patent Grant
11256696
Not Applicable.
Not Applicable.
This invention relates generally to computer networking and more particularly to database system and operation.
Computing devices are known to communicate data, process data, and/or store data. Such computing devices range from wireless smart phones, laptops, tablets, personal computers (PC), work stations, and video game devices, to data centers that support millions of web searches, stock trades, or on-line purchases every day. In general, a computing device includes a central processing unit (CPU), a memory system, user input/output interfaces, peripheral device interfaces, and an interconnecting bus structure.
As is further known, a computer may effectively extend its CPU by using “cloud computing” to perform one or more computing functions (e.g., a service, an application, an algorithm, an arithmetic logic function, etc.) on behalf of the computer. Further, for large services, applications, and/or functions, cloud computing may be performed by multiple cloud computing resources in a distributed manner to improve the response time for completion of the service, application, and/or function.
Of the many applications a computer can perform, a database system is one of the largest and most complex applications. In general, a database system stores a large amount of data in a particular way for subsequent processing. In some situations, the hardware of the computer is a limiting factor regarding the speed at which a database system can process a particular function. In some other instances, the way in which the data is stored is a limiting factor regarding the speed of execution. In yet some other instances, restricted co-process options are a limiting factor regarding the speed of execution.
In an example of operation, the parallelized data input sub-system 11 receives tables of data from a data source. For example, a data set no. 1 is received when the data source includes one or more computers. As another example, the data source is a plurality of machines. As yet another example, the data source is a plurality of data mining algorithms operating on one or more computers. The data source organizes its data into a table that includes rows and columns. The columns represent fields of data for the rows. Each row corresponds to a record of data. For example, a table include payroll information for a company's employees. Each row is an employee's payroll record. The columns include data fields for employee name, address, department, annual salary, tax deduction information, direct deposit information, etc.
The parallelized data input sub-system 11 processes a table to determine how to store it. For example, the parallelized data input sub-system 11 divides the data into a plurality of data partitions. For each data partition, the parallelized data input sub-system 11 determines a number of data segments based on a desired encoding scheme. As a specific example, when a 4 of 5 encoding scheme is used (meaning any 4 of 5 encoded data elements can be used to recover the data), the parallelized data input sub-system 11 divides a data partition into 5 segments. The parallelized data input sub-system 11 then divides a data segment into data slabs. Using one or more of the columns as a key, or keys, the parallelized data input sub-system 11 sorts the data slabs. The sorted data slabs are sent, via the system communication resources 14, to the parallelized data store, retrieve, and/or process sub-system 12 for storage.
The parallelized query and response sub-system 13 receives queries regarding tables and processes the queries prior to sending them to the parallelized data store, retrieve, and/or process sub-system 12 for processing. For example, the parallelized query and response sub-system 13 receives a specific query no. 1 regarding the data set no. 1 (e.g., a specific table). The query is in a standard query format such as Open Database Connectivity (ODBC), Java Database Connectivity (JDBC), and/or SPARK. The query is assigned to a node within the sub-system 13 for subsequent processing. The assigned node identifies the relevant table, determines where and how it is stored, and determines available nodes within the parallelized data store, retrieve, and/or process sub-system 12 for processing the query.
In addition, the assigned node parses the query to create an abstract syntax tree. As a specific example, the assigned node converts an SQL (Standard Query Language) statement into a database instruction set. The assigned node then validates the abstract syntax tree. If not valid, the assigned node generates a SQL exception, determines an appropriate correction, and repeats. When the abstract syntax tree is validated, the assigned node then creates an annotated abstract syntax tree. The annotated abstract syntax tree includes the verified abstract syntax tree plus annotations regarding column names, data type(s), data aggregation or not, correlation or not, subquery or not, and so on.
The assigned node then creates an initial query plan from the annotated abstract syntax tree. The assigned node optimizes the initial query plan using a cost analysis function (e.g., processing time, processing resources, etc.). Once the query plan is optimized, it is sent, via the system communication resources 14, to the parallelized data store, retrieve, and/or process sub-system 12 for processing.
Within the parallelized data store, retrieve, and/or process sub-system 12, a computing device is designated as a primary device for the query plan and receives it. The primary device processes the query plan to identify nodes within the parallelized data store, retrieve, and/or process sub-system 12 for processing the query plan. The primary device then sends appropriate portions of the query plan to the identified nodes for execution. The primary device receives responses from the identified nodes and processes them in accordance with the query plan. The primary device provides the resulting response to the assigned node of the parallelized query and response sub-system 13. The assigned node determines whether further processing is needed on the resulting response (e.g., joining, filtering, etc.). If not, the assigned node outputs the resulting response as the response to the query (e.g., a response for query no. 1 regarding data set no. 1). If, however, further processing is determined, the assigned node further processes the resulting response to produce the response to the query.
As will be described in greater detail with reference to one or more subsequent figures, a computing device includes a plurality of nodes and each node includes a plurality of processing core resources. Each processing core resource is capable of executing at least a portion of an administrative operation independently. This supports lock free and parallel execution of one or more administrative operations.
As will be described in greater detail with reference to one or more subsequent figures, a computing device includes a plurality of nodes and each node includes a plurality of processing core resources. Each processing core resource is capable of executing at least a portion of an configuration operation independently. This supports lock free and parallel execution of one or more configuration operations.
The parallelized ingress sub-system 24 includes a plurality of ingress data sub-systems 25-1 through 25-p that each include a local communication resource of local communication resources 26-1 through 26-p and a plurality of computing devices 18-1 through 18-n. Each of the computing devices of the parallelized ingress sub-system 24 execute an ingress data processing function utilizing an ingress data processing of ingress data processing 28-1 through 28-n of each ingress data sub-system 25-1 through 25-p that enables the computing device to stream data of a table (e.g., a data set 30-2 as segments 29-1-1 through 29-1-n and through 29-1-p through 29-n-p) into the database system 10 of
Each of the bulk data processing function and the ingress data processing function generally function as described with reference to
As will be described in greater detail with reference to one or more subsequent figures, a computing device includes a plurality of nodes and each node includes a plurality of processing core resources. Each processing core resource is capable of executing at least a portion of the bulk data processing function or the ingress data processing function. In an embodiment, a plurality of processing core resources of one or more nodes executes the bulk data processing function or the ingress data processing function to produce the storage format for the data of a table.
The Q & R function enables the computing devices to processing queries and create responses as discussed with reference to
As will be described in greater detail with reference to one or more subsequent figures, a computing device includes a plurality of nodes and each node includes a plurality of processing core resources. Each processing core resource is capable of executing at least a portion of the IO & P function. In an embodiment, a plurality of processing core resources of one or more nodes executes the IO & P function to produce at least a portion of the resulting response as discussed in
In an embodiment, each node is capable of operating independently of the other nodes. This allows for large scale parallel operation of a query request, which significantly reduces processing time for such queries. In another embodiment, one or more node function as co-processors to share processing requirements of a particular function, or functions.
The disk memory 38 includes a plurality of memory interface modules 43-1 through 43-n and a plurality of memory devices 42-1 through 42-n. The memory devices 42-1 through 42-n include, but are not limited to, solid state memory, disk drive memory, cloud storage memory, and other non-volatile memory. For each type of memory device, a different memory interface module 43-1 through 43-n is used. For example, solid state memory uses a standard, or serial, ATA (SATA), variation, or extension thereof, as its memory interface. As another example, disk drive memory devices use a small computer system interface (SCSI), variation, or extension thereof, as its memory interface.
In an embodiment, the disk memory 38 includes a plurality of solid state memory devices and corresponding memory interface modules. In another embodiment, the disk memory 38 includes a plurality of solid state memory devices, a plurality of disk memories, and corresponding memory interface modules.
The network connection 41 includes a plurality of network interface modules 46-1 through 46-n and a plurality of network cards 47-1 through 47-n. A network card includes a wireless LAN (WLAN) device (e.g., an IEEE 802.11n or another protocol), a LAN device (e.g., Ethernet), a cellular device (e.g., CDMA), etc. The corresponding network interface modules 46-1 through 46-n include a software driver for the corresponding network card and a physical connection that couples the network card to the central processing module 39 or other component(s) of the node.
The connections between the central processing module 39, the main memory 40, the disk memory 38, and the network connection 41 may be implemented in a variety of ways. For example, the connections are made through a node controller (e.g., a local version of the computing device controller hub 36). As another example, the connections are made through the computing device controller hub 36.
The main memory 40 is divided into a computing device (CD) 56 section and a database (DB) 51 section. The database section includes a database operating system (OS) area 52, a disk area 53, a network area 54, and a general area 55. The computing device section includes a computing device operating system (OS) area 57 and a general area 58. Note that each section could include more or less allocated areas for various tasks being executed by the database system.
In general, the database OS 52 allocates main memory for database operations. Once allocated, the computing device OS 57 cannot access that portion of the main memory 40. This supports lock free and independent parallel execution of one or more operations.
The database overriding operating system (DB OS) 61 includes custom DB device management 69, custom DB process management 70 (e.g., process scheduling and/or inter-process communication & synchronization), custom DB file system management 71, custom DB memory management 72, and/or custom security 73. In general, the database overriding OS 61 provides hardware components of a node for more direct access to memory, more direct access to a network connection, improved independency, improved data storage, improved data retrieval, and/or improved data processing than the computing device OS.
In an example of operation, the database overriding OS 61 controls which operating system, or portions thereof, operate with each node and/or computing device controller hub of a computing device (e.g., via OS select 75-1 through 75-n when communicating with nodes 37-1 through 37-n and via OS select 75-m when communicating with the computing device controller hub 36). For example, device management of a node is supported by the computer operating system, while process management, memory management, and file system management are supported by the database overriding operating system. To override the computer OS, the database overriding OS provides instructions to the computer OS regarding which management tasks will be controlled by the database overriding OS. The database overriding OS also provides notification to the computer OS as to which sections of the main memory it is reserving exclusively for one or more database functions, operations, and/or tasks. One or more examples of the database overriding operating system are provided in subsequent figures.
The parallelized query & response sub-system is capable of receiving and processing a plurality of queries in parallel. For ease of discussion, the present method is discussed with reference to one query.
The method branches to steps 145 and 151. At step 145, the computing device identifies a table (or tables) for the received query. The method continues at step 147 where the computing device determines where and how the table(s) is/are stored. For example, the computing device determines how the table was partitioned; how each partition was divided into one or more segment groups; how many segments in a segment group; how many storage clusters are storing segment groups; how many computing devices are in a storage cluster; how many nodes per computing device; and/or how many processing core resources per node.
The method continues at step 149 where the computing device determines available nodes (and/or processing core resources) within the parallelized Q&R sub-system for processing operations of the query. In addition, the computing device determines nodes (and/or processing core resources) available for processing operations of the query. Typically, the nodes and/or processing core resources storing a relevant portion of the table will be need for processing one or more operations of the query.
At step 151, the computing device parses the received query to create an abstract syntax tree. For example, the computing device converts SQL statements of the query into nodes of a syntactic structure of source code and creates a tree structure of the nodes. A node corresponds to a construct occurring in the source code.
The method continues at step 153 where the computing device validates the abstract syntax tree. For example, the computing device verifies one or more of the SQL statements are valid, the conversion to operations of the DB instruction set are valid, the table(s) exists, the selected operations of the DB instruction set and/or the SQL statements yield viable data (e.g., will produce a result, will not cause a deadlock, etc.), etc. If not, the computing device sends an SQL exception to the source of the query.
For validated abstract syntax tree, the method continues at step 155 where the computing device generates an annotated abstract syntax tree. For example, the computing device adds column names, data types, aggregation information, correlation information, subquery information, etc. to the verified abstract system tree.
The method continues at step 157 where the computing device creates an initial query plan from the annotated abstract syntax tree. For example, the computing device selects operations from an operating instruction set of the database system to implement the abstract syntax tree. The operating instruction set of the database system (i.e., DB instruction set) includes the following operations:
Aggregation—aggregates two or more rows based on one or more values of a row and then combine (e.g., sum, average, appended, sort, etc.) into a row;
AggVectorOperationInstance—use when number of rows is known and is less than or equal to a specific value (e.g., 256), use a vector operation instead of a hash function to aggregate rows, which allows aggregation without the need for caching;
Broadcast—computing device or node sending data to other computing devices or nodes performing similar tasks, functions, and/or operations (typically for lateral data flow in the system);
Eos—“end of stream” is a placeholder to indicate no data, may also be used to indicate a function cannot be performed;
Except—set subtraction;
Extend—add a column to received data;
Gather—combine data together;
GdeLookup—“Global Dictionary Compression” lookup function for data compression;
HashJoin—join data using a hash function;
IncrementBigInt—increment one or more data values in accordance with a test protocol
IncremetingInt—increment one or more data values
Index—uses indexed metadata to reduce amount of data to read and/or to push operations downstream to delay reading;
IndexAgg—aggregation of indexing;
IndexDistinct—indexing of distinct row, rows, column, and/or columns;
SegmentAgg (operator instance)—segmenting of an aggregation operation to produce sub-aggregation operations;
SegmentDistinct (operator instance)—segmenting of a distinct operation to produce sub-distinct operations;
IndexCountStar
Intersect—is a mathematical function to find data from two or more sets of data that intersect;
JobsVirtual
Limit—limit the number of rows to be read, to be operated on, etc.;
MakeVector—convert columns into a matrix for linear algebra functions;
UnMakeVector—convert a resulting matrix back into columns;
MatrixExtend—add columns or another matrix to an existing matrix;
Offset—is an offset for data retrieval;
OrderedAgg—ordering of aggregation to allow for lower level aggregation, which allows higher level to be more efficient;
OrderedDistinct—ordering of distinct values at lower levels, which allows higher levels to be more efficient;
OrderedGather—ordering of gathering at lower levels, which allows higher levels to be more efficient;
ProductJoin—nested loop join function (e.g., join data from one or more rows and/or from one or more columns);
ProjectOut—remove a column for data of interest (e.g., want to do this as far downstream as possible);
Rename—change name of a column, (can be used to avoid column name collisions);
Reorder—reorder data of one or more rows and/or one or more columns based on an ordering preference;
Root—conduit for data flow;
Select—select columns from one or more tables;
Shuffle—sub-divide data into a plurality of data sub-divisions (typically for lateral data flow in the system);
Switch—change where to send data when a condition is met;
TableScan—retrieve all of the data of a table;
TableSlab Scan (operator instance)—retrieve particular data slabs of a table;
Tee—creates a brand in operational flow when operating on redundant data;
Union—establish a set of operations;
Window—is a specific type of aggregation that captures a moving window of aggregated data (e.g., a running sum, a running average, etc.); and
MultiplexerOperatorInstance for Set/ProductJoin/HashJoin/Sort/Aggregation—allows for lock free multiplexing for various types of operations.
The method continues at step 159 where the computing device optimizes the query plan using a cost analysis of step 161. The initial query plan is created to be executed by a computing device within the parallelized query & response sub-system. Optimizing the plan spreads the execution of the query across multiple layers (e.g., three or more) and to include the other sub-systems of the database system. The computing device utilizes one or more optimization transforms to optimize the initial query plan. The optimization transforms include:
AddDistinctBeforeMinMax: Adds a union distinct before an aggregation operator that only performs min/max
RemoveDistinctBeforeMinMax: The opposite of addDistinctBeforeMinMax
AddDistinctBetoreSemiAnti: Adds a union distinct as the right child of a join that is a semi or anti join
RemoveDistinctBeforeSemiAnti: The opposite of addDistinctBeforeSemiAnti
AggDistinctPushDown: Pushes down an aggregation that is only performing distinct operators (count/sum distinct) below its child
AggDistinctPushUp: The opposite of AggDistinctPushDown
AggregatePushDown: The same as AggDistinctPushDown but for aggregations performing non-distinct operations
AggregatePushUp: The opposite of AggregatePushDown
ConvertProductToHashJoin: Converts a product join with lhasCol=rhsCol filters into an equivalent hash join
CreateTee: Given a certain node in the tree, searches the rest of the tree for equivalent subtrees, if one or more is found, the equivalent subtrees are deleted and a tee operator is created as the parent of the given node, which then forwards the results to the parents of those equivalent subtrees
Delete Tee: The opposite of create Tee
RedistributeAggDistinct: Moves a distinct aggregation to a lower level (below a gather), and adds a shuffle if needed
DedistributeAggDistinct: The opposite of redistributeAggDistinct
RedistibuteAggregation: The same as redistributeAggDistinct but for non-distinct aggregations
DedistributeAggregation: The opposite of redistributeAggregation
DeletePointlessSort: Deletes a pointless sort from the tree
DeletePointlessSwitch: Deletes a pointless switch from the tree (only happens if all of the extends the switch created were pushed out of the switch-union block)
DuplicateAggBelowShuffles: Given an aggregation (including aggdistinct) with a shuffle as its child, create a copy of the aggregation below the shuffle and update the original to have the correct operations
RemoveAggBelowShuffles: The opposite of duplicateAggBelowShuffles
DuplicateLimit: Given a limit above a gather type operator, create a copy of it below the gather type operator
ExceptPushDown: Pushes an except operator down below all of its child, can only happen if they are all equivalent
ExceptPushUp: The opposite of exceptPushDown
ExceptUnionContract: Given an except with more than 2 children, take children [1, N−1] and make them the children of a union all, which becomes child 1 of the except
ExceptUnionExpand: The opposite of exceptUnionContract
ExtendPushDown
ExtendPushUp
IntersectPushDown: The same as exceptPushDown but for an intersect operator
IntersectPushUp: The opposite of intersectPushDown
JoinPushDown: Pushes a join down below its child(ren). Similar to except/intersectPushDown except with a few other cases. If one child is a join it instead swaps the joins, it also has to check that pushing below its children doesn't break the join (for example by creating name collisions or removing columns that needed to exist)
JoinPushUp: The opposite of joinPushDown, but with some more potential for optimizations. Specifically, if the parent is a select on equiJoin columns, the select can be pushed down to all children, or is the parent is a project and the join is a gdcJoin, then this deletes the join and its right subtree entirely
LimitPushDown
LimitPushUp
MakeVectorDown
MakeVectorPushUp
MatrixExtendPushDown
MatrixExtendPushI)own
MergeEquiJoins: Given two adjacent inner hash joins with no other filters, combine them into a single hash join with more children
SplitEquiJoins: The opposite of mergeEquiJoins
MergeExcept: Given two adjacent except operators, take the input to the lower one and make all of its children become children of the higher one
MergeIntersect: The same as mergeExcept but for intersect
MergeTee: Given two adjacent tee operators, take delete the higher one and make its parent additional parents on the lower one
MergeUnion: The same as mergeExcept but for union
MergeWindows: Combine two adjacent window operators into a single one
OffsetPushDown
OffsetPushUp
ProjectOutPushDown
ProjectOutPushUp
PushAggBelowJoin: Duplicates an aggregation below a hash join, and updates the higher one accordingly
PushAggAboveJoin: The opposite of pushAggBelowJoin
PushAggBelowGdcJoin: Given an aggregation above a gdcJoin, this moves it below the gdcJoin if possible. Currently requires that the aggregation does not reference the gdc column at all, or only groups by it. More cases are possible
PushJoinBelowSet: Given a join where one if it's children is a set operator, and moves the join below the set such that there are not multiple joins as the children of the set operator
PushSetBelowJoin: The opposite of pushJoinBelowSet
PushLimitintoIndex: Pushes a limit operator into an index operator, this way the index knows to only output up to LIMIT rows
PushLimitIntoSort: Pushes a limit into a sort operator, which causes us to run a faster limitSort algorithm in the virtual machine (e.g., node or processing core resource)
PushLimitOutOfSort: The opposite of pushLimitIntoSort
PushProjectIntoIndex: Pushes a project into an operator, which causes a not read of a column. Used when start reading all columns in plan generation
PushSelectBelowGdcJoin: Given a select above a gdcJoin, where the select is filtering the compressed column, this converts the filter to a filter on the stored integer mapping of that column, and moves the select below the join. For example, where coll=“hello” might be converted to where coll Key=42
PushSelectintoHashJoin: Given a select above a hash join, where the select filters on lhsCol=rhsCol, this creates additional equi join columns on the hash join
PushSelectOutOffiashJoin: The opposite of pushSelectintoHashJoin
PushSelectintoProduct: The same as pushSelectintoHashJoin but for product joins
PushSelectOut01Product: The opposite of pushSelectIntoProduct
RenamePushDown
RenamePushUp
ReorderPushDown
ReorderPushUp
SelectOutJoinNulls: Given a join that is joining on coil, if coil is nullable this creates a select below the join that has the filter where coil !=NULL
UnselectOutJoinNulls: The opposite of selectOutJoinNulls
SelectPushDown
SelectPushUp
SortPushDown
SortPushUp
SwapJoinChildren: Swaps the order of a joins children
SwitchPushDown: Given a switch operator, push it down over its child. In some cases, this causes copies of the child to become the switch's parents', and in others this causes that child to jump the entire switch union block and become the parent of the union associated with the switch
SwitchPushUp: The opposite of switchPushDown, but nothing jumps because the parents of the switch are inside the switch union block already. Also requires that all parents are equivalent
TeePushDown: Pushes a tee down below its child, causing that child to be copied for each parent of the tee
TeePushUp: The opposite of teePushDown, requires that all parents are equivalent
UnionDistinctCopyDown: Given a union distinct with gathers as its children, creates another 1 child union distinct as the children of those gathers
UnionDistinctCopyUp: The opposite of unionDistinctCopyDown
UnionPushDown: The same as exceptPushDown except for union, also handles the different rules that apply to union all and union distinct
UnionPushlJp: The opposite of unionPushDown, also handles the case where this is the opposite of switchPushDown because the union has an associated switch, so some operators will jump the entire switch union block
UnmakeVectorPushDown
UnmakeVectorPushUp
WindowPushDown
WindowPushUp
post-optimization options
Combining adjacent selects into super Selects
Combining adjacent limits
Combining adjacent offsets
Converting distinct aggregations into a non-distinct aggregation with a union distinct as its child
Duplicating union distincts around shuffles, this only happens if there is a union distinct on 1 side of a shuffle, but not both
Replacing index type operators with an eos operator we if can determine that the filters (if any) on the index are always false (possible by comparing possible values of data types)
Evaluating alternate indexes besides the primary index
Building orderedAggregations and orderedDistincts
Getting rid of pointless renames
Pushing sorts down to level 3 if possible
Creating indexCountStar operators if possible
Fixing out of order indexAggs, this makes the grouping key order match the primary index order when possible
Tee'ing leaf operators, this combines as many equivalent leaf operators as possible to reduce IO
Deleting pointless reorders
Note that the Down and push Up transforms are used frequently, and mean to take the given operator and swap its position in the tree with its child (or parent) for most operators. Further note that not all of these transforms are legal in all possible cases, and they only get applied if they are legal.
The method continues at step 163 where the query plan is executed to produce a query result. The execution of the query plan is discussed in greater detail in subsequent figures.
For example, four threads of operations include a multiplex sort. The downstream operations in the threads (e.g., the operations on the bottom of the figure) execution an operation to produce a result or data value. For each result or data value that falls in range “a” is sent upstream to the operation in the far-left thread. For each result or data value that falls in range “b” is sent upstream to the operation in the second from the left thread. For each result or data value that falls in range “c” is sent upstream to the operation in the second from the right thread. For each result or data value that falls in range “d” is sent upstream to the operation in the far-right thread.
The operations use a bucket sort operation when the results or data values are of a defined set of values (e.g., integers, dates, time, etc.) to identify the appropriate upstream operation. When the results or data values are not of defined set of values (e.g., names, floating point data, etc.), the operations use a normal sort function to identify the appropriate upstream operation.
As a specific example, assume that range “a” is from negative infinity to −1 million; range “b” is from −999,999 to −1; range “c” is from 0 to 999,999; and range “d” is from +1 million to infinity. As such, the downstream operations would use one or more normal sort functions for ranges “a” and “d” and uses one or more bucket sort functions for ranges “b” and “c”.
If, at step 203, the data set has at least some known possible values, the method continues at step 207 where the processing core resource determines whether the lowest range is bounded. For example, when there is a specific lowest value (e.g., −1 million), then the lowest range is bounded. As another example, when there is not a specific lowest value (e.g., −infinity), the lowest range is not bounded. When the lowest range is not bounded, the method continues at step 209 where the processing core resource uses a normal sort function for the lowest range. Whether the lowest range is bounded or not, the method continues at step 211 where the processing core resource determines whether the highest range is bounded. If not, the method continues at step 213 where the processing core resources uses a normal sort function for the highest range. Whether or not the highest range is bounded, the method continues at step 215 where the processing core resource uses a bucket sort function for all other ranges that have not yet been flagged for a normal sort function.
Accordingly, when a data block is written into the disk memory section 53 of the database (DB) memory space 51, it is done so as a data block with each data word having a sequential address. This facilitates direct memory access of the main memory 40 by the memory devices via the respective memory interfaces.
Data messages includes a corresponding message address of message addresses 230-1, 230-2 etc. and a plurality of data blocks 232-1 through 232-n. Each data block has an associated block address of block addresses 234-1 through 234-n. The block addresses are logical addressees and are sequential within a data message. The message address corresponds to the first data block address and the other data block addresses are a logical offset from the first. For example, a data message is 1 M Byte in size and includes 256 4 Kbyte data blocks. This message data structure within the DB network section 54 of the main memory 40 facilitates the network connection to have direct memory access.
The selection of the L2 computing devices 18 can be done in a variety of ways. For example, the L2 computing device is selected based on a pseudo random selection process. As another example, the L2 computing device is selected in a round-robin manner. Having selected the L2 computing devices for each storage cluster, the computing device 18 of the parallelized data input sub-system 11 sends a corresponding segment group of segment groups 1_1 through 1_n to each L2 computing device.
Each LBA includes a number of fixed size data fields 240 positioned within the LBA. In an example LBAi through LBA+x includes 27 (128) positions and each block of data includes 4,096 positions. In practice, the number of positions, data value, and data fields can be any reasonable value. In the example of
The administrative sub-system 15 creates global dictionary compression (GDC) 246 tables based on requests 242 from the parallelized data input sub-system 11 and/or based on requests 244 from the parallelized data, store, retrieve, and/or process sub-system 12. For example, a request includes a request for the administrative sub-system 15 to create or update a city dictionary. As another example, a request includes a request for the administrative sub-system 15 to create or update a state dictionary.
In a second example of implementing the global dictionary compression, the parallelized data input sub-system 11 receives a data set (e.g., one or more tables 236) that includes a plurality of data records. Each data record of the plurality of data records includes a plurality of data fields. A data record of the plurality of data records includes a first data field of the plurality of data fields containing a first fixed length data value of a plurality of fixed length data values (e.g., record numbers, SSN, employee number, etc.) and a second data field of the plurality of data fields containing a first variable length data value of a plurality of variable length data values (e.g., names, city, state, etc.).
The data set has a first organizational structure. The first organizational structure of the data set includes one of a first table format where rows of a first table are the data records and columns of the first table are the data fields, a second table format where the columns of a second table are the data records and the rows of the second table are the data fields, and a tree structure where the data records are linked in a hierarchical order. The first variable length data value includes one or more of a binary string that represents one of: text data, audio data, video data, image data, graphics data, and numerical data, and an alpha-numeric string that represents one of: text data, audio data, video data, image data, graphics data, and numerical data.
Having received the data set, the parallelized data input sub-system 11 accesses (e.g., utilizing the request 242 to the administrative sub-system 15 and receiving the dictionary 246 in response) a compression dictionary for the second data field. The compression dictionary includes a plurality of entries, where each entry of the plurality of entries includes a key field and a value field. A first entry of the plurality of entries includes the key field storing a first fixed length index value and the value field storing the first variable length data value of the plurality of variable length data values. The key field has a smaller data size than the value field.
The accessing the compression dictionary includes determining, by the parallelized data input sub-system 11, whether the compression dictionary for the second data field exists. When the compression dictionary for the second data field does not exist, the parallelized data input sub-system 11 initiates creation of the compression dictionary for the second data field (e.g., generates the dictionary and/or sends the request 242 to the administrative sub-system 15 and receives the dictionary 246 in response). When the compression dictionary does exist, the parallelized data input sub-system 11 accesses the compression dictionary (e.g., in a local memory). When creating the compression dictionary and/or updating the compression dictionary, the parallelized data input sub-system 11 updates the compression dictionary with a new entry for a new variable length data value being added to the plurality of variable length data values.
Having accessed the compression dictionary, the parallelized data input sub-system 11 creates a storage data set based on the data set and the compression dictionary, where the first variable length data value of the second data field of the data record is replaced with the first fixed length index value. The storage data set has a plurality of fixed length fields. The creating the storage data set further includes one or more of replacing a second variable length data value of the second data field of a second data record of the plurality of data records of the data set with a corresponding second fixed length index value of a second entry of the plurality of entries of the compression dictionary (e.g., a different record with different variable length value), and replacing the first variable length data value of the second data field of a third data record of the plurality of data records of the data set with the first fixed length index value (e.g., a different record with same variable length value).
When a third data field is required, the parallelized data input sub-system 11 may access a second compression dictionary for the third data field of the plurality of data fields, where the second compression dictionary includes a second plurality of entries, where each entry of the second plurality of entries includes a second key field and a second value field. A first entry of the second plurality of entries includes the second key field storing a second fixed length index value and the second value field storing a second variable length data value of a second plurality of variable length data values, where the second key field has a smaller data size than the second value field.
Having accessed the second compression dictionary for the third data field, the parallelized data input sub-system 11 creates the storage data set based on the data set, the compression dictionary, and the second compression dictionary, where the second variable length data value of the third data field of the data record is replaced with the second fixed length index value. The creating the storage data set further includes selecting the first data field of the data set, selecting the value field from the compression dictionary, selecting the second value field from the second compression dictionary, joining the data set to the compression dictionary based on the first data field of the data set and the value field of the compression dictionary, joining the data set to the second compression dictionary based on the first data field of the data set and the second value field of the second compression dictionary, and creating a view name for the storage data set that corresponds to a name of the data set.
When the storage data set has been created, the parallelized data input sub-system 11 sends the storage data set to a data storage-process sub-system for storage. For example, the parallelized data input sub-system 11 sends the storage data set as segments for storage 241 to the parallelized data store, retrieve, &/or process sub-system 12 for storage.
To mimic the user's table, but taking advantage of global dictionary compression, the administration sub-system creates a new table (SYSDDC.USER.TABLE), which is designated as table 1. Table 1 includes three columns (C0, C1, and C2), but each are integer columns. Column C1 includes integers that are keys into a second table (e.g., SYSLOOKUP.USER.TABLE_C1). The second table includes two columns. The first is an integer column that includes the keys or codes for the string values of the user's table in column 1 (e.g., cities).
Column C2 of the new table includes integers that are keys into a third table (e.g., SYSLOOKUP.USER.TABLE_C2). The third table includes two columns. The first is an integer column that includes the keys or codes for the string values of the user's table in column 2 (e.g., states).
If the data was stored using GDC, then the method continues at step 267 where the processing node identifies an operation, or operations, of the initial plan that has a GDC data operand(s) (e.g., is access data that was compressed using GDC). The method continues at step 269 where the processing node determines whether the operation itself, or a sequence of operations, can be optimized (e.g., reworked to more efficiently access data and/or more efficiently process data). If yes, the method continues at step 271 where the processing node optimizes the operation and/or the sequence of operations.
Whether the operation or sequence of operations are optimized or not, the method continues at step 273 where the processing node determines whether the operation, or sequence of operations can be performed without GDC decoding. For example, if the operation or sequence of operations is to count the records by state, the name of the state is not needed for this operation. As such, decoding is not needed. If yes, the method continues at step 281 where the processing node optimizes the operation to use the GDC code without GDC decoding.
If, however, the operation cannot be performed without GDC decoding (e.g., adding floating point values of a list of floating point values), the method continues at step 275 where the processing node determines whether the operation needs to done at the current level or can the operation be pushed upstream. If the operation can be pushed upstream, the method continues at step 277 where the processing node moves the operation upstream.
When the operation cannot be pushed upstream, or pushed upstream any further, the method continues at step 279 where the processing node inserts a GDC join operation to executed the GDC decoding, which replaces the key code with the actual value. The method continues at step 283 where the processing node determines whether the plan optimization is complete. If so, the method ends. If not, the method repeats at step 267 for another operation, or sequence of operations, that access data that has been compressed using GDC.
The method continues at step 302 where the processing module accesses a compression dictionary for the second data field, where the compression dictionary includes a plurality of entries, and where each entry of the plurality of entries includes a key field and a value field. A first entry of the plurality of entries includes the key field storing a first fixed length index value and the value field storing the first variable length data value of the plurality of variable length data values. The key field has a smaller data size than the value field. The accessing the compression dictionary includes determining whether the compression dictionary for the second data field exists and when the compression dictionary for the second data field does not exist, initiating creation of the compression dictionary for the second data field (e.g., creating the dictionary or obtaining the compression dictionary from another computing entity of the data processing system). When the compression dictionary does exist, the processing module accesses the compression dictionary. When a new entry is to be processed, the processing module updates the compression dictionary with the new entry for a new variable length data value being added to the plurality of variable length data values.
The method continues at step 304 where the processing module creates a storage data set based on the data set and the compression dictionary, where the first variable length data value of the second data field of the data record is replaced with the first fixed length index value, and where the storage data set has a plurality of fixed length fields. The creating the storage data set further includes one or more of replacing a second variable length data value of the second data field of a second data record of the plurality of data records of the data set with a corresponding second fixed length index value of a second entry of the plurality of entries of the compression dictionary (e.g., a different record with different variable length value), and replacing the first variable length data value of the second data field of a third data record of the plurality of data records of the data set with the first fixed length index value (e.g., a different record with a same variable length value).
The method continues at step 306 when operating on the third data field, otherwise the method continues to step 310. When operating on the third data field, the processing module accesses a second compression dictionary for a third data field of the plurality of data fields, where the second compression dictionary includes a second plurality of entries. Each entry of the second plurality of entries includes a second key field and a second value field. A first entry of the second plurality of entries includes the second key field storing a second fixed length index value and the second value field storing a second variable length data value of a second plurality of variable length data values, where the second key field has a smaller data size than the second value field.
The method continues at step 308 where the processing module creates the storage data set based on the data set, the compression dictionary, and the second compression dictionary, where the second variable length data value of the third data field of the data record is replaced with the second fixed length index value. The creating the storage data set further includes selecting the first data field of the data set, selecting the value field from the compression dictionary, selecting the second value field from the second compression dictionary, joining the data set to the compression dictionary based on the first data field of the data set and the value field of the compression dictionary, joining the data set to the second compression dictionary based on the first data field of the data set and the second value field of the second compression dictionary, and creating a view name for the storage data set that corresponds to a name of the data set.
When the storage data set has been created, the method continues at step 310 where the processing module sends the storage data set to a data storage-process sub-system of the data processing system for storage. For example, the processing module sends the storage data set to the data storage-process sub-system for direct storage. In another example, the processing module sends the storage data set to the data storage-process sub-system for further compression optimization and storage, where the further compression optimization includes utilizing one or more of the compression dictionary, the second compression dictionary, and another compression dictionary.
The method described above in conjunction with the processing module can alternatively be performed by other modules of the database system 10 of
It is noted that terminologies as may be used herein such as bit stream, stream, signal sequence, etc. (or their equivalents) have been used interchangeably to describe digital information whose content corresponds to any of a number of desired types (e.g., data, video, speech, audio, etc. any of which may generally be referred to as ‘data’).
As may be used herein, the terms “substantially” and “approximately” provides an industry-accepted tolerance for its corresponding term and/or relativity between items. Such an industry-accepted tolerance ranges from less than one percent to fifty percent and corresponds to, but is not limited to, component values, integrated circuit process variations, temperature variations, rise and fall times, and/or thermal noise. Such relativity between items ranges from a difference of a few percent to magnitude differences. As may also be used herein, the term(s) “configured to”, “operably coupled to”, “coupled to”, and/or “coupling” includes direct coupling between items and/or indirect coupling between items via an intervening item (e.g., an item includes, but is not limited to, a component, an element, a circuit, and/or a module) where, for an example of indirect coupling, the intervening item does not modify the information of a signal but may adjust its current level, voltage level, and/or power level. As may further be used herein, inferred coupling (i.e., where one element is coupled to another element by inference) includes direct and indirect coupling between two items in the same manner as “coupled to”. As may even further be used herein, the term “configured to”, “operable to”, “coupled to”, or “operably coupled to” indicates that an item includes one or more of power connections, input(s), output(s), etc., to perform, when activated, one or more its corresponding functions and may further include inferred coupling to one or more other items. As may still further be used herein, the term “associated with”, includes direct and/or indirect coupling of separate items and/or one item being embedded within another item.
As may be used herein, the term “compares favorably”, indicates that a comparison between two or more items, signals, etc., provides a desired relationship. For example, when the desired relationship is that signal 1 has a greater magnitude than signal 2, a favorable comparison may be achieved when the magnitude of signal 1 is greater than that of signal 2 or when the magnitude of signal 2 is less than that of signal 1. As may be used herein, the term “compares unfavorably”, indicates that a comparison between two or more items, signals, etc., fails to provide the desired relationship.
As may be used herein, one or more claims may include, in a specific form of this generic form, the phrase “at least one of a, b, and c” or of this generic form “at least one of a, b, or c”, with more or less elements than “a”, “b”, and “c”. In either phrasing, the phrases are to be interpreted identically. In particular, “at least one of a, b, and c” is equivalent to “at least one of a, b, or c” and shall mean a, b, and/or c. As an example, it means: “a” only, “b” only, “c” only, “a” and “b”, “a” and “c”, “b” and “c”, and/or “a”, “b”, and “c”.
As may also be used herein, the terms “processing module”, “processing circuit”, “processor”, and/or “processing unit” may be a single processing device or a plurality of processing devices. Such a processing device may be a microprocessor, micro-controller, digital signal processor, microcomputer, central processing unit, field programmable gate array, programmable logic device, state machine, logic circuitry, analog circuitry, digital circuitry, and/or any device that manipulates signals (analog and/or digital) based on hard coding of the circuitry and/or operational instructions. The processing module, module, processing circuit, and/or processing unit may be, or further include, memory and/or an integrated memory element, which may be a single memory device, a plurality of memory devices, and/or embedded circuitry of another processing module, module, processing circuit, and/or processing unit. Such a memory device may be a read-only memory, random access memory, volatile memory, non-volatile memory, static memory, dynamic memory, flash memory, cache memory, and/or any device that stores digital information. Note that if the processing module, module, processing circuit, and/or processing unit includes more than one processing device, the processing devices may be centrally located (e.g., directly coupled together via a wired and/or wireless bus structure) or may be distributedly located (e.g., cloud computing via indirect coupling via a local area network and/or a wide area network). Further note that if the processing module, module, processing circuit, and/or processing unit implements one or more of its functions via a state machine, analog circuitry, digital circuitry, and/or logic circuitry, the memory and/or memory element storing the corresponding operational instructions may be embedded within, or external to, the circuitry comprising the state machine, analog circuitry, digital circuitry, and/or logic circuitry. Still further note that, the memory element may store, and the processing module, module, processing circuit, and/or processing unit executes, hard coded and/or operational instructions corresponding to at least some of the steps and/or functions illustrated in one or more of the Figures. Such a memory device or memory element can be included in an article of manufacture.
One or more embodiments have been described above with the aid of method steps illustrating the performance of specified functions and relationships thereof. The boundaries and sequence of these functional building blocks and method steps have been arbitrarily defined herein for convenience of description. Alternate boundaries and sequences can be defined so long as the specified functions and relationships are appropriately performed. Any such alternate boundaries or sequences are thus within the scope and spirit of the claims. Further, the boundaries of these functional building blocks have been arbitrarily defined for convenience of description. Alternate boundaries could be defined as long as the certain significant functions are appropriately performed. Similarly, flow diagram blocks may also have been arbitrarily defined herein to illustrate certain significant functionality.
To the extent used, the flow diagram block boundaries and sequence could have been defined otherwise and still perform the certain significant functionality. Such alternate definitions of both functional building blocks and flow diagram blocks and sequences are thus within the scope and spirit of the claims. One of average skill in the art will also recognize that the functional building blocks, and other illustrative blocks, modules and components herein, can be implemented as illustrated or by discrete components, application specific integrated circuits, processors executing appropriate software and the like or any combination thereof.
In addition, a flow diagram may include a “start” and/or “continue” indication. The “start” and “continue” indications reflect that the steps presented can optionally be incorporated in or otherwise used in conjunction with other routines. In this context, “start” indicates the beginning of the first step presented and may be preceded by other activities not specifically shown. Further, the “continue” indication reflects that the steps presented may be performed multiple times and/or may be succeeded by other activities not specifically shown. Further, while a flow diagram indicates a particular ordering of steps, other orderings are likewise possible provided that the principles of causality are maintained.
The one or more embodiments are used herein to illustrate one or more aspects, one or more features, one or more concepts, and/or one or more examples. A physical embodiment of an apparatus, an article of manufacture, a machine, and/or of a process may include one or more of the aspects, features, concepts, examples, etc. described with reference to one or more of the embodiments discussed herein. Further, from figure to figure, the embodiments may incorporate the same or similarly named functions, steps, modules, etc. that may use the same or different reference numbers and, as such, the functions, steps, modules, etc. may be the same or similar functions, steps, modules, etc. or different ones.
Unless specifically stated to the contra, signals to, from, and/or between elements in a figure of any of the figures presented herein may be analog or digital, continuous time or discrete time, and single-ended or differential. For instance, if a signal path is shown as a single-ended path, it also represents a differential signal path. Similarly, if a signal path is shown as a differential path, it also represents a single-ended signal path. While one or more particular architectures are described herein, other architectures can likewise be implemented that use one or more data buses not expressly shown, direct connectivity between elements, and/or indirect coupling between other elements as recognized by one of average skill in the art.
The term “module” is used in the description of one or more of the embodiments. A module implements one or more functions via a device such as a processor or other processing device or other hardware that may include or operate in association with a memory that stores operational instructions. A module may operate independently and/or in conjunction with software and/or firmware. As also used herein, a module may contain one or more sub-modules, each of which may be one or more modules.
As may further be used herein, a computer readable memory includes one or more memory elements. A memory element may be a separate memory device, multiple memory devices, or a set of memory locations within a memory device. Such a memory device may be a read-only memory, random access memory, volatile memory, non-volatile memory, static memory, dynamic memory, flash memory, cache memory, and/or any device that stores digital information. The memory device may be in a form a solid-state memory, a hard drive memory, cloud memory, thumb drive, server memory, computing device memory, and/or other physical medium for storing digital information.
While particular combinations of various functions and features of the one or more embodiments have been expressly described herein, other combinations of these features and functions are likewise possible. The present disclosure is not limited by the particular examples disclosed herein and expressly incorporates these other combinations.
The present U.S. Utility Patent Application claims priority pursuant to 35 U.S.C. § 119(e) to U.S. Provisional Application No. 62/745,787, entitled “DATABASE SYSTEM AND OPERATION,” filed Oct. 15, 2018, which is hereby incorporated herein by reference in its entirety and made part of the present U.S. Utility Patent Application for all purposes.
| Number | Name | Date | Kind |
|---|---|---|---|
| 5548770 | Bridges | Aug 1996 | A |
| 6230200 | Forecast | May 2001 | B1 |
| 6633772 | Ford | Oct 2003 | B2 |
| 7499907 | Brown | Mar 2009 | B2 |
| 7908242 | Achanta | Mar 2011 | B1 |
| 20010051949 | Carey | Dec 2001 | A1 |
| 20020032676 | Reiner | Mar 2002 | A1 |
| 20040162853 | Brodersen | Aug 2004 | A1 |
| 20080133456 | Richards | Jun 2008 | A1 |
| 20090006399 | Raman | Jan 2009 | A1 |
| 20090063893 | Bagepalli | Mar 2009 | A1 |
| 20090183167 | Kupferschmidt | Jul 2009 | A1 |
| 20100082577 | Mirchandani | Apr 2010 | A1 |
| 20100241646 | Friedman | Sep 2010 | A1 |
| 20100274983 | Murphy | Oct 2010 | A1 |
| 20100312756 | Zhang | Dec 2010 | A1 |
| 20100328115 | Binnig | Dec 2010 | A1 |
| 20110219169 | Zhang | Sep 2011 | A1 |
| 20120109888 | Zhang | May 2012 | A1 |
| 20120151118 | Flynn | Jun 2012 | A1 |
| 20120185866 | Couvee | Jul 2012 | A1 |
| 20120254252 | Jin | Oct 2012 | A1 |
| 20120311246 | Mcwilliams | Dec 2012 | A1 |
| 20130332484 | Gajic | Dec 2013 | A1 |
| 20140047095 | Breternitz | Feb 2014 | A1 |
| 20140136510 | Parkkinen | May 2014 | A1 |
| 20140188841 | Sun | Jul 2014 | A1 |
| 20150205607 | Lindholm | Jul 2015 | A1 |
| 20150244804 | Warfield | Aug 2015 | A1 |
| 20150248366 | Bergsten | Sep 2015 | A1 |
| 20150293966 | Cai | Oct 2015 | A1 |
| 20150310045 | Konik | Oct 2015 | A1 |
| 20160034547 | Lerios | Feb 2016 | A1 |
| 20160378545 | Ho | Dec 2016 | A1 |
| Entry |
|---|
| A new high performance fabric for HPC, Michael Feldman, May 2016, Intersect360 Research. |
| Alechina, N. (2006-2007) B-Trees School of Computer Science, University of Nottingham, http://www.cs.nott.ac.uk/˜psznza/G5BADS06/lecture13-print.pdf. 41 pages. |
| Amazon DynamoDB: ten things you really should know, Nov. 13, 2015, Chandan Patra, http://cloudacademy.com/blog/amazon-dynamodb-ten-thing. |
| An Inside Look at Google BigQuery, by Kazunori Sato, Solutions Architect, Cloud Solutions team, Google Inc., 2012. |
| Big Table, a NoSQL massively parallel table, Paul Krzyzanowski, Nov. 2011, https://www.cs.rutgers.edu/pxk/417/notes/contentibigtable.html. |
| Distributed Systems, Fall2012, Mohsen Taheriyan, http://www-scf.usc.edu/-csci57212011Spring/presentations/Taheriyan.pptx. |
| International Searching Authority; International Search Report and Written Opinion; International Application No. PCT/US2017/054773; dated Feb. 13, 2018; 17 pgs. |
| International Searching Authority; International Search Report and Written Opinion; International Application No. PCT/US2017/054784; dated Dec. 28, 2017; 10 pgs. |
| International Searching Authority; International Search Report and Written Opinion; International Application No. PCT/US2017/066145; dated Mar. 5, 2018; 13 pgs. |
| International Searching Authority; International Search Report and Written Opinion; International Application No. PCT/US2017/066169; dated Mar. 6, 2018; 15 pgs. |
| International Searching Authority; International Search Report and Written Opinion; International Application No. PCT/US2018/025729; dated Jun. 27, 2018; 9 pgs. |
| International Searching Authority; International Search Report and Written Opinion; International Application No. PCT/US2018/034859; dated Oct. 30, 2018; 8 pgs. |
| MapReduce: Simplied Data Processing on Large Clusters, OSDI2004, Jeffrey Dean and Sanjay Ghemawat, Google, Inc. |
| Rodero-Merino, L.; Storage of Structured Data: Big Table and HBase, New Trends In Distributed Systems, MSc Software and Systems, Distributed Systems Laboratory; Oct. 17, 2012; 24 pages. |
| Step 2: Examine the data model and implementation details, 2016, Amazon Web Services, Inc., http://docs.aws.amazon.com/amazondynamodb/latestldeveloperguide!Ti . . . . |
| Number | Date | Country | |
|---|---|---|---|
| 20200117649 A1 | Apr 2020 | US |
| Number | Date | Country | |
|---|---|---|---|
| 62745787 | Oct 2018 | US |
