Presto Verifier

Presto Verifier is a tool to run queries and verify correctness. It can be used to test whether a new Presto version produces the correct query results, or to test if pairs of Presto queries have the same semantics.

During each Presto release, Verifier is run to ensure that there is no correctness regression.

Using Verifier

In a MySQL database, create the following table and load it with the queries you would like to run:

  1. CREATE TABLE verifier_queries (
  2. id int(11) unsigned NOT NULL PRIMARY KEY AUTO_INCREMENT,
  3. suite varchar(256) NOT NULL,
  4. name varchar(256) DEFAULT NULL,
  5. control_catalog varchar(256) NOT NULL,
  6. control_schema varchar(256) NOT NULL,
  7. control_query text NOT NULL,
  8. control_username varchar(256) DEFAULT NULL,
  9. control_password varchar(256) DEFAULT NULL,
  10. control_session_properties text DEFAULT NULL,
  11. test_catalog varchar(256) NOT NULL,
  12. test_schema varchar(256) NOT NULL,
  13. test_query text NOT NULL,
  14. test_username varchar(256) DEFAULT NULL,
  15. test_password varchar(256) DEFAULT NULL,
  16. test_session_properties text DEFAULT NULL)

Next, create a config.properties file:

  1. source-query.suites=suite
  2. source-query.database=jdbc:mysql://localhost:3306/mydb?user=my_username&password=my_password
  3. control.hosts=127.0.0.1
  4. control.http-port=8080
  5. control.jdbc-port=8080
  6. control.application-name=verifier-test
  7. test.hosts=127.0.0.1
  8. test.http-port=8081
  9. test.jdbc-port=8081
  10. test.application-name=verifier-test
  11. test-id=1

Verifier can run in either query-bank or control-test mode by setting configuration running-mode.

  • control-test: this is the default mode. Both the control query and the test query are executed and their result checksums are compared.

  • query-bank: in this mode, the control query is skipped and the comparison is done between a saved snapshot result and the test result.

Create a verifier_snapshots table:

  1. CREATE TABLE verifier_snapshots (
  2. id int(11) unsigned NOT NULL PRIMARY KEY AUTO_INCREMENT,
  3. suite varchar(256) NOT NULL,
  4. name varchar(256) NOT NULL DEFAULT '.',
  5. is_explain BOOLEAN NOT NULL DEFAULT false,
  6. snapshot json NOT NULL,
  7. updated_at datetime NOT NULL DEFAULT now(),
  8. UNIQUE(suite, name, is_explain));

Download presto-verifier-0.289-executable.jar and rename it to verifier.jar. To run the Verifier:

  1. java -Xmx1G -jar verifier.jar verify config.properties

Before running in query-bank mode, snapshots must be saved. Add configurations:

  1. running-mode=query-bank
  2. save-snapshot=true

Run the verifier and the snapshots will be saved to the table verifier_snapshots.

To run in query-bank mode, set save-snapshot=false or just delete it:

  1. running-mode=query-bank
  2. #save-snapshot=true

Verifier Procedures

The following steps summarize the workflow of Verifier.

  • Importing Source Queries

    • Reads the list of source queries (query pairs with configuration) from the MySQL table.
  • Query Pre-processing and Filtering

    • Applies overrides to the catalog, schema, username, and password of each query.

    • Filters queries according to whitelist and blacklist. Whitelist is applied before blacklist.

    • Filters out queries with invalid syntax.

    • Filters out queries not supported for validation. Select, Insert, CreateTableAsSelect, create table and create view are supported.

  • Query rewriting

    • Rewrites queries before execution to ensure that production data is not modified.

    • Rewrites Select queries to CreateTableAsSelect

      • Column names are determined by running the Select query with LIMIT 0.

      • Artificial names are used for unnamed columns.

  1. - Rewrites `Insert` and `CreateTableAsSelect` queries to have their table names replaced.
  2. - Constructs a setup query to create the table necessary for an `Insert` query.
  3. - Rewrites function calls according to `nondeterministic-function-substitutes` if the configuration is set.
  • Query Execution

    • Conceptually, Verifier is configured with a control cluster and a test cluster. However, they may be pointed to the same Presto cluster for certain tests. In Query-bank mode, the control cluster is skipped and a saved snapshot is used instead.

    • For each source query, executes the following queries in order.

      • Control setup queries

      • Control query

      • Test setup queries

      • Test query

      • Control and test teardown queries

  1. - Queries are subject to timeouts and retries.
  2. - Cluster connection failures and transient Presto failures are retried.
  3. - Query retries may conceal reliability issues, and therefore Verifier records all occurred Presto query failures, including the retries.
  4. - Certain query failures are automatically submitted for re-validation, such as partition dropped or table dropped during query.
  5. - See [Failure Resolution](#failure-resolution) for auto-resolving of query failures.
  • Results Comparison

    • For Select, Insert, and CreateTableAsSelect queries, results are written into temporary tables.

    • Constructs and runs the checksum queries for both control and test. If running in query-bank mode, Control checksums will be restored from snapshots saved in mysql database.

    • Control checksums will be saved to mysql database if configuration save-snapshot is set to true.

    • Verifies table schema and row count are the same for the control and the test result table.

    • Verifies checksums are matching for each column. See Column Checksums for special handling of different column types.

    • See Determinism for handling of non-deterministic queries.

  • Emitting Results

    • Verification results can be exported as JSON, or human readable text.

Column Checksums

For each column in the control/test query, one or more columns are generated in the checksum queries.

  • Floating Point Columns

    • For DOUBLE and REAL columns, 4 columns are generated for verification:

      • Sum of the finite values of the column

      • NAN count of the column

      • Positive infinity count of the column

      • Negative infinity count of the column

    • Checks if NAN count, positive and negative infinity count matches.

    • Checks the nullity of control sum and test sum.

    • If either control mean or test mean very close 0, checks if both are close to 0.

    • Checks the relative error between control sum and test sum.

  • VARCHAR Columns

    • For VARCHAR columns, a simple checksum column is generated for verification using the checksum().

    • If validate-string-as-double is set, seven columns below are generated. If NULL counts are equal before and after casting all values to DOUBLE, apply the floating point validation. Otherwise, check if the simple checksum matches.

      • Checksum

      • Count of the NULL values

      • Count of the NULL values after all values are casted to DOUBLE

      After casting all values to DOUBLE

      • Sum of the finite values

      • NAN count of the values

      • Positive infinity count of the values

      • Negative infinity count of the values

  • Array Columns

    • For an array column arr of type array(E), three columns are generated for verification:

      • Sum of the cardinality

      • Checksum of the cardinality

      • Array checksum

        • If E is not orderable, array checksum is checksum(arr).

        • If E is orderable, array checksum is coalesce(checksum(try(array_sort(arr))), checksum(arr)).

  1. - If `use-error-margin-for-floating-point-arrays` is set and E is `DOUBLE` or `REAL`, these six columns below are generated instead. Check if the sum of the cardinality matches and if the checksum of the cardinality matches. Apply the floating point validation to the rest of results.
  2. - Sum of the cardinality
  3. - Checksum of the cardinality
  4. - Sum of the finite elements of all array values
  5. - Count of the `NAN` elements of all array values
  6. - Count of the positive infinity elements of all array values
  7. - Count of the negative infinity elements of all array values
  8. - If `validate-string-as-double` is set and E is `VARCHAR`, these nine columns below are generated instead. Check if the sum and the checksum of the cardinality match. If `NULL` counts are equal before and after casting all array elements to `DOUBLE`, apply the floating point validation. Otherwise, check if the array checksum matches.
  9. - Sum of the cardinality
  10. - Checksum of the cardinality
  11. - Array checksum `checksum(array_sort(arr))`
  12. - Count of the `NULL` elements of all array values
  13. - Count of the `NULL` elements after all array elements are casted to `DOUBLE`
  14. After casting all array elements to `DOUBLE`
  15. - Sum of the finite elements of all array values
  16. - Count of the `NAN` elements of all array values
  17. - Count of the positive infinity elements of all array values
  18. - Count of the negative infinity elements of all array values
  • Map Columns

    • For a map column of type map(K, V), four columns are generated for verification:

      • Sum of the cardinality

      • Checksum of the map

      • Array checksum of the key set

      • Array checksum of the value set

  1. - If `validate-string-as-double` is set and K is `VARCHAR`, six additional columns are generated:
  2. - Count of the `NULL` elements of all key sets
  3. - Count of the `NULL` elements of the key sets after all map keys are casted to `DOUBLE`
  4. After casting all map keys to `DOUBLE`
  5. - Sum of the finite elements of all key sets
  6. - Count of the `NAN` elements of all key sets
  7. - Count of the positive infinity elements of all key sets
  8. - Count of the negative infinity elements of all key sets
  9. - If `validate-string-as-double` is set and V is `VARCHAR`, six additional columns are generated:
  10. - Count of the `NULL` elements of all value sets
  11. - Count of the `NULL` elements of the value sets after all map values are casted to `DOUBLE`
  12. After casting all map values to `DOUBLE`
  13. - Sum of the finite elements of all value sets
  14. - Count of the `NAN` elements of all value sets
  15. - Count of the positive infinity elements of all value sets
  16. - Count of the negative infinity elements of all value sets
  • Row Columns

    • Checksums row fields recursively according to the type of the fields.

For all other column types, generates a simple checksum using the checksum() function.

Determinism

A result mismatch, either a row count mismatch or a column mismatch, can be caused by non-deterministic query features. To avoid false alerts, we perform determinism analysis for the control query. If a query is found non-deterministic, we skip the verification as it does not provide insights.

Determinism analysis follows the following steps. If a query is found non-deterministic at any point, the analysis will conclude.

  • Non-deterministic catalogs can be specified with determinism.non-deterministic-catalog. If a query references any table from those catalogs, the query is considered non-deterministic.

  • Runs the control query again and compares the results with the initial control query run.

  • If a query has a LIMIT n clause but no ORDER BY clause at the top level:

    • Runs a query to count the number of rows produced by the control query without the LIMIT clause.

    • If the resulting row count is greater than n, treats the control query as non-deterministic.

Failure Resolution

The differences in configuration, including cluster size, can cause a query to succeed on the control cluster but fail on the test cluster. A checksum query can also fail, which may be due to limitation of Presto or Presto Verifier. Thus, we allow Verifier to automatically resolve certain query failures.

  • EXCEEDED_GLOBAL_MEMORY_LIMIT: Resolves if the control query uses more memory than the test query.

  • EXCEEDED_TIME_LIMIT: Resolves unconditionally.

  • TOO_MANY_HIVE_PARTITIONS: Resolves if the test cluster does not have enough workers to make sure the number of partitions assigned to each worker stays within the limit.

  • COMPILER_ERROR, GENERATED_BYTECODE_TOO_LARGE: Resolves if the control checksum query fails with this error. If the control query has too many columns, generated checksum queries might be too large in certain cases.

In cases of result mismatches, Verifier may be giving noisy signals, and we allow Verifier to automatically resolve certain mismatches.

  • Structured-typed Columns: If array element or map key/value contains floating point types, column checksum is unlikely to match.

    • For an array column, resolve if the element type contains floating point types and the cardinality checksum matches.

    • For a map column, resolve the mismatch when both of the following conditions are true:

      • The cardinality checksum matches.

      • The checksum of the key or value that does not contains floating point types matches.

    • Resolve a test case only when all columns are resolved.

  • Resolved Functions: In the case of a results mismatch, if the query uses a function in a

    specified list, the test case is marked as resolved.

Explain Mode

In explain mode, Verifier checks whether source queries can be explained instead of whether they produces the same results. Verification is marked as succeeded when both control query and test query can be explained.

The field matchType in the output event can be used as an indicator whether there are plan differences between the control run and the test run.

For non-DML queries, the control query and the plan comparison are skipped.

Extending Verifier

Verifier can be extended for further behavioral changes in addition to configuration properties.

AbstractVerifyCommand shows the components that be extended. Implement the abstract class and create a command line wrapper similar to PrestoVerifier.

Configuration Reference

General Configuration

Name

Description

whitelist

A comma-separated list specifying the names of the queries within the suite to verify.

blacklist

A comma-separated list specifying the names of the queries to be excluded from the suite. blacklist is applied after whitelist.

source-query-supplier

The name of the source query supplier. Supports mysql.

source-query.table-name

The name of the table that holds verifier queries. Available only when source-query-supplier is mysql.

event-clients

A comma-separated list specifying where the output events should be emitted. Supports json and human-readable.

json.log-file

The output files of JSON events. If not set, JSON events are emitted to stdout.

human-readable.log-file

The output files for human-readable events. If not set, human-readable events are emitted to stdout.

control.table-prefix

The table prefix to be appended to the control target table.

test.table-prefix

The table prefix to be appended to the test target table.

control.reuse-table

If true, reuse the output table of the control source Insert and CreateTableAsSelect query. Otherwise, run the control source query and write to a temporary table.

test.reuse-table

If true, reuse the output table of the test source Insert and CreateTableAsSelect query. Otherwise, run the test source query and write to a temporary table.

test-id

A string to be attached to output events.

max-concurrency

Maximum number of concurrent verifications.

suite-repetition

How many times a suite is verified.

query-repetition

How many times a source query is verified.

relative-error-margin

Maximum tolerable relative error between control sum and test sum of a floating point column.

absolute-error-margin

Floating point averages that are below this threshold are treated as 0.

run-teardown-on-result-mismatch

Whether to run teardown query in case of result mismatch.

verification-resubmission.limit

A limit on how many times a source query can be re-submitted for verification.

running-mode

Set to query-bank to make the Verifier run in query-bank mode. Supports query-bank and control-test. Defaults to control-test.

save-snapshot

Set to true to save checksums to mysql database.

extended-verification

Set to true to run extended table layout verification for written tables. It only applies to Insert and CreateTableAsSelect queries. It would verify each partition’s data checksum if the inserted table is partitioned. It would verify each bucket’s data checksum if the inserted table is bucketed.

function-substitutes

Specification of function substitutions, in the format of /foo(c0,)/bar(c0)/,/fred(c0,c1)/baz(qux(c1,c0))/,/foobar(c0)/if(qux(c1),bar(c0),baz(c1))/,…, where foo(c0, ) would be substituted by bar(c0), with the declared arguments applied to the corresponding positions.

Concatenate function substitutions with a comma.

Select a function substitute that has the return type and argument types compatible with those of the original function, to produce a valid source query. For example, /array_agg(z)/array_sort(array_agg(z))/,/approx_percentile(x,y)/avg(x)/.

Declare the function arguments as identifiers if they need to be applied to the function substitute.

Query Override Configuration

The following configurations control the behavior of query metadata modification before verification starts. Counterparts are also available for test queries with prefix control being replaced with test.

Name

Description

control.catalog-override

The catalog to be applied to all queries if specified.

control.schema-override

The schema to be applied to all queries if specified.

control.username-override

The username to be applied to all queries if specified.

control.password-override

The password to be applied to all queries if specified.

control.session-properties-override-strategy

Supports 3 values. NO_ACTION: Use the session properties as specified for each query. OVERRIDE: Merge the session properties of each query with the override, with override being the dominant. SUBSTITUTE, The session properties of each query is replaced with the override.

control.session-properties-override

The session property to be applied to all queries.

Query Execution Configuration

The following configurations control the behavior of query execution on the control cluster. Counterparts are also available for test clusters with prefix control being replaced with test.

Name

Description

control.hosts

Comma-separated list of the control cluster hostnames or IP addresses.

control.jdbc-port

JDBC port of the control cluster.

control.http-host

HTTP port of the control cluster.

control.jdbc-url-parameters

A JSON map representing the additional URL parameters for control JDBC.

control.query-timeout

The execution time limit of the control and the test queries.

control.metadata-timeout

The execution time limit of DESC queries and LIMIT 0 queries.

control.checksum-timeout

The execution time limit of checksum queries.

control.application-name

ApplicationName to be passed in ClientInfo. Can be used to set source.

Determinism Analyzer Configuration

Name

Description

determinism.run-teardown

Whether to run teardown queries for tables produced in determinism analysis.

determinism.max-analysis-runs

Maximum number of additional control runs to check for the determinism of the control query.

determinism.handle-limit-query

Whether to enable the special handling for queries with a top level LIMIT clause.

determinism.non-deterministic-catalogs

A comma-separated list of non-deterministic catalogs. Queries referencing table from those catalogs are treated as non-deterministic.

Failure Resolution Configuration

Name

Description

exceeded-global-memory-limit.failure-resolver.enabled

Whether to enable the failure resolver for test query failures with EXCEEDED_GLOBAL_MEMORY_LIMIT.

exceeded-time-limit.failure-resolver.enabled

Whether to enable the failure resolver for test query failures with EXCEEDED_TIME_LIMIT.

verifier-limitation.failure-resolver.enabled

Whether to enable the failure resolver for failures due to Verifier limitations.

too-many-open-partitions.failure-resolver.enabled

Whether to enable the failure resolver for test query failures with HIVE_TOO_MANY_OPEN_PARTITIONS.

too-many-open-partitions.max-buckets-per-writer

The maximum buckets count per writer configured on the control and the test cluster.

too-many-open-partitions.cluster-size-expiration

The time limit of the test cluster size being cached.

structured-column.failure-resolver.enabled

Whether to enable the failure resolver for column mismatches of structured-type columns.

ignored-functions.failure-resolver.enabled

Whether to enable the IgnoredFunctions result mismatch failure resolver.

ignored-functions.functions

A comma-separated list of functions. Resolves mismatches if a query uses any functions in the list.