yoklainterview sim

Data Engineer Wh Bigquery Storage Partitioning Clustering Interview Questions

75 verified Data Engineer Wh Bigquery Storage Partitioning Clustering interview questions — solve with answers, learn from explanations, test yourself in a real simulation.

Try the real simulation →

Sample questions

Wh Bigquery Storage Partitioning ClusteringDifficulty 1
Which three strategies can you use to partition a table in BigQuery?
  • aHash key, range key, and composite key
  • bPrimary key, foreign key, and surrogate key
  • cTime-unit, ingestion time, integer range
  • dRow-level, column-level, and block-level
Explanation:BigQuery offers three partitioning strategies: a time-unit column (DATE/TIMESTAMP/DATETIME), ingestion time (based on load time), and integer range on an INTEGER column.
Wh Bigquery Storage Partitioning ClusteringDifficulty 1
For a table using ingestion-time partitioning with no explicit partitioning column, which pseudocolumns does BigQuery expose to reference the partition?
  • a_PARTITIONTIME and _PARTITIONDATE
  • b_INGESTED_AT and _LOAD_DATE
  • c_ROW_TIMESTAMP and _PARTITION_KEY
  • d_EVENT_TIME and _BATCH_ID
Explanation:Ingestion-time partitioned tables expose the pseudocolumns _PARTITIONTIME (TIMESTAMP) and _PARTITIONDATE (DATE) so queries can filter on the load-time partition.
Wh Bigquery Storage Partitioning ClusteringDifficulty 1
What three values do you specify to define the ranges of an integer-range partitioned table?
  • aA minimum value, a maximum value, and a partition count
  • bA start value, an end, and an interval
  • cA seed value, a hash bucket count, and a salt
  • dA base value, a growth rate, and a ceiling
Explanation:Integer-range partitioning is defined by a start (inclusive), an end (exclusive), and an interval, e.g. RANGE_BUCKET(customer_id, GENERATE_ARRAY(0, 100, 10)).
Wh Bigquery Storage Partitioning ClusteringDifficulty 1
At most how many columns can you specify when clustering a BigQuery table?
  • aTwo
  • bEight
  • cSixteen
  • dFour
Explanation:BigQuery allows at most four clustering columns per table, specified with CLUSTER BY.
Wh Bigquery Storage Partitioning ClusteringDifficulty 2
Why does the order of columns matter when you define clustering on a BigQuery table?
  • aA filter benefits only when it covers a leading prefix of the clustered columns
  • bThe first clustered column determines which region the table's storage lives in
  • cColumn order only changes how the table looks in the console UI, not query cost
  • dBigQuery re-sorts the given columns alphabetically regardless of the order you list
Explanation:Clustering benefit is prefix-based: a query filtering on the first N clustered columns (in the declared order) can prune blocks, but filtering only on a later column skips the benefit of the earlier ones.
Wh Bigquery Storage Partitioning ClusteringDifficulty 2
What does BigQuery actually do when you cluster a table by one or more columns?
  • aIt builds a secondary B-tree index over the clustered columns
  • bIt moves the clustered columns into a separate, smaller physical table
  • cIt orders storage blocks by the clustered values, so matching queries scan less
  • dIt caches the clustered columns in memory for the duration of the session
Explanation:Clustering physically co-locates and sorts rows into storage blocks by the clustered column values, so a query filtering or aggregating on those columns only needs to read the relevant blocks.

Test yourself against the 1950-question Data Engineer bank.

Start interview