Cluster Computing Functions

Read(3752) Label: function, node,

  esProc offers cluster computing functions that handle cluster computations.

Ø  callx(spl,…;hs;rdfx)

Allocate tasks among a sequence of nodes, execute the script, and return a sequence of values returned by all tasks

Ø  cs.fetch()

Fetch data from a cluster cursor

Ø  cs.group()

Group a cluster cursor by comparing each record with its next neighbor

Ø  cs.group(x:F,...;y:G,…)

Group a cluster cursor by comparing each record with its next neighbor and then aggregate each group

Ø  cs.groups()

Group records in a cluster cursor

Ø  cs.groupx()

Group records in a cluster cursor and return result as a synchronously segmented cluster cursor

Ø  cs.join()

A foreign-key-style join between a cluster cursor and a record sequence

Ø  cs.memory()

Generate a cluster in-memory table from a cluster cursor

Ø  cs.sortx()

Sort data in a cluster cursor

Ø  cs.switch()

Replace values of a specified field in a cluster cursor with the referencing field values of the corresponding record in another table

Ø  file(fn:cs,h)

Return a remote file from a node

Ø  file(fn:cs,hs)

Return a remote file from a sequence of nodes

Ø  file(fn:z,f)

Generate a new cluster homo-name files group with same distribution as an existing one

Ø  file(fn:z,hs)

Generate a cluster file on one or more nodes

Ø  fork ….;hs ,reduce

Execute a code block in a node

Ø  f.open()

Generate a cluster table from a cluster file

Ø  hosts(i;j)

Set or get the number of memory zone holding a specified task on the current node

Ø  hosts(n,hs;j)

Return a sequence of available nodes

Ø  joinx()

Join multiple synchronously segmented cluster cursors/multicursors

Ø  memory()

Generate a cluster in-memory table from a local in-memory table

Ø  movefile(fn,h;path,hs)

Move a file from a node to another

Ø  splserver()

Start or stop an SPL process on the current machine

Ø  syncfile(hs,p)

Synchronize a local path onto a sequence of nodes

Ø  T.attach(T’)

Retrieve an attached table from a cluster composite table

Ø  T.close()

Close a cluster table

Ø  T.cursor(x,…;wi,...)

Return one or more specified columns of a cluster entity table/cluster in-memory table as a cluster cursor

Ø  T.cursor(x:C,…;wi,...;mcs)

Synchronously segment a cluster entity table according to a cluster multicursor and return a cluster multicursor

Ø  T.dup()

Join up a cluster in-memory table into a local in-memory table

Ø  T.dup(h)

Copy a local in-memory table onto nodes to create a duplicate cluster in-memory table

Ø  T.memory(C,;w)

Generate a cluster in-memory table from a cluster entity table