---

Parallel algorithm for σ_cond(R)

---

• Recall: the uni-processor algorithm
  • One-pass algorithm:

    initialize a search structure H on all attributes of R; while ( R has more data blocks ) { read 1 data block in buffer b; for ( each tuple t ∈ b ) { /* ===================================================== We need a search structure H to implement the test t ∈ H efficiently !!! We can use hash table or some bin. search tree ====================================================== */ if ( t ∈ H ) { discard t // duplicate } else { insert t in H; // Help find duplicates move t to output // This is the first occurence } } }

    Buffer utilization when there are M buffers available:

  ---

  ---

  ---

  ---

• Parallel algorithm for δ(R)
  • Parallel algorithm for δ(R):

    Each processor executes the single-processor δ( ) locally on its fragment of relation R:

    Result:

    The collection (bag) of the output of (at) each processor is the output of δ(R)

  ---

  ---

  ---

  ---

• Why it works
  • Duplicates can be eliminated locally because:

    duplicate tuples will always stored at the same processor !!!

  ---

  ---

  ---

  ---

• Performance of the parallel δ(R)
  • Performance of the parallel δ(R):

    Distribution of the relation R: Each processor will perform: 1 --- B(R) disk block read operations P

    Therefore:

    The parallel δ(R) operation will complete in 1/P of the time