validate

Validate cluster object

Since R2026a

Syntax

validate(myCluster)

validate(___,Name=Value)

Description

validate(myCluster) validates the cluster specified by the cluster object myCluster. You can use the validate function to validate a cluster object after you modify a cluster object property.

The validate function prints the validation stages performed and an indication of whether each stage passed, failed, or was skipped.

If a validation stage fails, the function saves a validation report to the temporary folder.

validate(___,Name=Value) specifies additional cluster validation options using one or more name-value arguments.

Examples

Validate Cluster Object

Create a cluster object for a remote cluster with the profile name myMJSCluster, using the parcluster function.

c = parcluster(myMJSCluster);

Validate the cluster object, c. The validate function prints the validation stages performed and an indication of whether each stage passed, failed, or was skipped.

validate(c)

Beginning cluster validation.
Cluster connection test (parcluster)
   Stage started at 15:52:05.
   Using existing cluster object.
   ✓ Network latency: 6 ms
   ✓ Upload speed: 22915 KB/s
   ✓ Download speed: 43480 KB/s
   For more information on these results, see the
   "Troubleshooting Slow Network Connection" documentation
   page.
   Finished at 15:52:11.
......................................................PASSED
Job test (createJob)
   Stage started at 15:52:11.
   Finished at 15:52:21.
......................................................PASSED
SPMD job test (createCommunicatingJob)
   Stage started at 15:52:21.
   Job ran with 1 worker.
   Finished at 15:52:26.
......................................................PASSED
Pool job test (createCommunicatingJob)
   Stage started at 15:52:26.
   Job ran with 1 worker.
   Finished at 15:52:45.
......................................................PASSED
Parallel pool test (parpool)
   Stage started at 15:52:45.
   Connected to parallel pool with 12 workers.
   Parallel pool is shutting down.
   Parallel pool ran with 12 worker.
   Finished at 15:52:53.
......................................................PASSED
Finished cluster validation with status: PASSED

Specify Stages to Validate

Create a cluster object using the default cluster profile.

c = parcluster;

Validate the cluster object, c using 5 workers. Run only the "job", "spmd-job", and "pool-job" stages. Save the validation results to a report file with the name "validation.txt".

validate(c,NumWorkersToUse=5, ...
    StagesToRun=["job","spmd-job","pool-job"],ReportFile="validation.txt")

Beginning cluster validation
Cluster connection test (parcluster)
   Stage started at 14:46:04.
   Using existing cluster object.
   Finished at 14:46:06.
..........................................................................PASSED

Job test (createJob)
   Stage started at 14:46:06.
   Finished at 14:46:52.
..........................................................................PASSED

SPMD job test (createCommunicatingJob)
   Stage started at 14:46:55.
   Job ran with 5 workers.
   Finished at 14:47:57.
..........................................................................PASSED

Pool job test (createCommunicatingJob)
   Stage started at 14:48:01.
   Job ran with 5 workers.
   Finished at 14:49:11.
..........................................................................PASSED

Finished validation with status: PASSED
Validation report written to validation.txt

Input Arguments

myCluster — Cluster

parallel.Cluster object

Cluster, specified as a parallel.Cluster object that represents cluster compute resources. To create the parallel.Cluster object, use the parcluster function.

Example: cluster = parcluster; validate(cluster);

Data Types: parallel.Cluster

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: NumWorkersToUse=4

NumWorkersToUse — Number of workers to use

nonnegative integer

Number of workers to use in validation, specified as a nonnegative integer. Use this argument to validate the cluster object with fewer workers than those specified in the cluster object's PreferredPoolNumWorkers or NumWorkers property.

Example: NumWorkersToUse=4

Data Types: single | double | int8 | int16 | int32 | int64 | uint8 | uint16 | uint32 | uint64

StagesToRun — Validation stages to run

string array | cell array of character vectors

Validation stages to run, specified as a string array or cell array of character vectors with one or more of these stage values:

• "parcluster" — Cluster connection test (parcluster) stage.

• "job" — Job test (createJob) stage.

• "spmd-job" — SPMD job test (createCommunicatingJob) stage.

• "pool-job" — Pool job test (createCommunicatingJob) stage.

• "parpool" — Parallel pool test (parpool) stage.

• "spark" — Spark job test (tall) stage.

• "tall" — Tall evaluation test (tall gather) stage.

Some stages are only available for certain parallel environment or cluster types. You cannot specify this argument if you specify the StagesToSkip argument. To learn more about each stage, see List of Validation Stages.

Example: StagesToRun=["job","parpool"]

Data Types: char | string | cell

StagesToSkip — Validation stages to skip

string array | cell array of character vectors

Validation stages to skip, specified as a string array or cell array of character vectors with one or more of these stage values:

• "parcluster" — Cluster connection test (parcluster) stage.

• "job" — Job test (createJob) stage.

• "spmd-job" — SPMD job test (createCommunicatingJob) stage.

• "pool-job" — Pool job test (createCommunicatingJob) stage.

• "parpool" — Parallel pool test (parpool) stage.

• "spark" — Spark job test (tall) stage.

• "tall" — Tall evaluation test (tall gather) stage.

Some stages are only available for certain parallel environment or cluster types. You cannot specify this argument if you specify the StagesToRun argument. To learn more about each stage, see List of Validation Stages.

Example: StagesToSkip=["job","spmd-job"]

Data Types: char | string | cell

ReportFile — File to save validation report

character vector | string

Filename to save validation report, specified as a character vector or string.

If you do not set the ReportFile argument and validation is successful, validate does not create a report. If you do not set the ReportFile argument and validation fails, validate automatically creates a validation report and saves it in the temporary folder.

Example: ReportFile="validation.txt"

Data Types: char | string

Algorithms

List of Validation Stages

The validate function performs these validation stages.

Validation Stage	Description
Cluster connection test (parcluster)	For remote cluster objects, tests cluster connections and displays cluster latency.
Job test (createJob)	Tests independent job creation using the createJob function.
SPMD job test (createCommunicatingJob)	Tests creation of a batch job with an SPMD enabled parallel pool using the createCommunicatingJob function.
Pool job test (createCommunicatingJob)	Tests creation of a batch job with a parallel pool using the createCommunicatingJob function.
Parallel pool test (parpool)	Tests creation of an interactive parallel pool using the parpool function.
Spark job test (tall)	Tests creation of a job on a Spark™ cluster.
Tall evaluation test (tall gather)	Tests evaluation of a tall array on a Spark cluster using the tall and gather functions.

Version History

Introduced in R2026a