Utilities | Boomerang Flow

The Utilities category has 7 tasks. Each entry shows the task's latest revision, the one the palette adds. See Reading an entry.

Execute Advanced HTTP Call

Execute an Advanced HTTP Call

Slug: execute-advanced-http-call
Type: template
Revision: 4
Image: Default task image.

Warning

Every revision sends no request and ends succeeded on boomerangio/task-flow:3.3.0, the default task image, so a later task that reads response or statusCode gets nothing. Fixed in task-flow 3.4.0: set Default Image Path to boomerangio/task-flow:3.4.0. See The HTTP Call tasks do nothing.

Parameters

Label	Name	Type	Required	Default	Helper text
URL	`url`	URL	No
Method	`method`	Select	No
Headers	`header`	Text Area	No		Headers to add to the request, such as Authorization
Content Type	`contentType`	Text	No
Body	`body`	Text Editor	No
Allow Untrusted SSL Certs	`allowUntrustedCerts`	Boolean	No
Response File Path	`outputFilePath`	Text	No		File path to store the output response
HTTP response codes for successful execution	`successcodes`	Text	No	`1xx,2xx`
HTTP response codes for failure	`errorcodes`	Text	No		HTTP response codes which marks the call as failed
HTTP response codes for execution retry	`retrycodes`	Text	No	`502,503`	HTTP response codes for retry
Number of milliseconds to wait until the next retry	`retrydelay`	Number	No		Cool down between retries in millis (100, 300000)
Number of retries	`retrynumber`	Number	No	`3`	Number of retries to be attempted, between (1, 9)

Results

Name	Description
`response`	HTTP execution response content
`statusCode`	The received HTTP status code

Example

Warning

This example works from task-flow 3.4.0: set Default Image Path to boomerangio/task-flow:3.4.0. On boomerangio/task-flow:3.3.0, the default task image, the task sends no request: see the warning above.

Record a deployment with an internal API, and retry while the service restarts.

Field	Value
Task name	`record-deploy`
URL	`https://deploy.example.com/api/deployments`
Method	`POST`
Headers	`Authorization: Bearer $(workspace.params.deploy-token)`
Content Type	`application/json`
Body	The JSON below
HTTP response codes for execution retry	Leave `502,503`
Number of retries	Leave `3`
Number of milliseconds to wait until the next retry	`2000`

Body

{"service": "$(params.service)", "version": "$(params.version)"}

The task returns:

Result	Value
`statusCode`	The status received, such as `201`
`response`	The response body as text, such as `{"id":"d-1042"}`, when it's 3,000 bytes or less

A 502 or 503 is retried up to 3 times, 2 seconds apart. Any 2xx succeeds. Any other status not in HTTP response codes for successful execution fails the task, with statusCode, and response when it fits, still set. Headers takes one Name: value per line, and only the names reach the log. For a larger response, set Response File Path to an absolute path or one relative to /workspace; response is then empty.

Common problems

Symptom	Cause	Fix
The task succeeds at once, and `statusCode` and `response` are empty	On `boomerangio/task-flow:3.3.0`, the default task image, the HTTP command is switched off.	Set Default Image Path to `boomerangio/task-flow:3.4.0`, or call the API with `curl --fail` in Execute Shell. See The HTTP Call tasks do nothing.
The task fails, saying the response is more than a task result can carry	The body is over 3,000 bytes, and Response File Path is blank.	Set Response File Path, and read the file in a later task.
A `301` or `302` fails the task	The task doesn't follow redirects.	Call the final URL, or add the code to HTTP response codes for successful execution.
The task fails before sending, saying the body isn't valid JSON	Content Type is a JSON type, and Body isn't valid JSON, often because a reference inserted a quote.	Correct the body. The task sends it as written.

Execute Basic HTTP Call

Execute a Basic HTTP Call

Slug: execute-basic-http-call
Type: template
Revision: 1
Image: Default task image.

Warning

Parameters

Label	Name	Type	Required	Default	Helper text
URL	`url`	URL	No
Method	`method`	Select	No
Headers	`header`	Text Area	No		Headers to add to the request, such as Authorizati
Content Type	`contentType`	Text	No
Body	`body`	Text Editor	No
Allow Untrusted SSL Certs	`allowUntrustedCerts`	Boolean	No	`false`
Response File Path	`outputFilePath`	Text	No		File path to store the output response

Results

Name	Description
`response`	HTTP execution response content
`statusCode`	The received HTTP status code

Execute Shell

Execute a shell script

Slug: execute-shell
Type: script
Revision: 2
Image: Default task image.

Parameters

Label	Name	Type	Required
Directory Path	`path`	Text	No
Shell Interpreter	`shell`	Text	No
Shell Script	`script`	Text Editor - Shell	Yes

No results.

Example

Make a file name from the workflow parameter version and return it as the result filename, for a later task to read. This is the build task in Pass results between tasks with an approval.

Field	Value
Task name	`build`
Directory Path	Blank, so the script runs in `/data`
Shell Interpreter	Blank, so BusyBox `sh` runs the script
Shell Script	The script below
Result Parameters	Name `filename`, Description `The file the build made`

Shell Script

filename="my-app-$(params.version).tar.gz"
echo "Built $filename"
if [ -d "$RESULTS_PATH" ]; then
  printf '%s' "$filename" > "$RESULTS_PATH/filename"
else
  printf '{"filename":"%s"}' "$filename" > "$RESULTS_PATH"
fi

The task returns:

Result	Value
`filename`	`my-app-1.4.2.tar.gz`, for a run started with `version` `1.4.2`

Flow replaces $(params.version) before the script runs, and a later task reads the result as $(tasks.build.results.filename). RESULTS_PATH is a directory on the tekton executor and a file on kube-jobs, and the if covers both. The default task image has bash, curl, sed, grep, tar and Node.js, but not jq. To use bash syntax, set Shell Interpreter to bash.

Common problems

Symptom	Cause	Fix
A later task reads an empty value from the result	The script didn't write the result to `$RESULTS_PATH`, or the node doesn't declare it under Result Parameters. Flow keeps only declared results.	Declare each result on the node, and write it as the example does. See Return results.
A later task can't find a file the script wrote	Directory Path is blank, so the script ran in `/data`, scratch space that only this task sees.	Turn on Shared workspace for each run under Configure › Workspace, then set Directory Path to `/workspace`, or write to paths under it.
The task fails with `No such file or directory: /workspace/`	On `boomerangio/task-flow:3.3.0`, Directory Path must be an absolute path that exists, with no trailing slash. A relative path, a trailing `/`, or `/workspace` without the run workspace each fail. From task-flow 3.4.0, a trailing slash or a path relative to `/workspace` works.	Enter an absolute path such as `/workspace`, or move to `boomerangio/task-flow:3.4.0`.
The task succeeds although a command in the script failed	Only the exit status of the script's last command counts.	Start the script with `set -e`, so the first command that fails ends the task Failed.

File JSONPath To Parameter

Takes the first value from the file content returned by a valid JSONPath expression and sets as an output property

Slug: file-jsonpath-to-parameter
Type: template
Revision: 2
Image: Default task image.

Parameters

Label	Name	Type	Required	Default	Helper text
File Path	`filePath`	Text	Yes		File content to query
Query Expression	`query`	Text	No		A valid JSON path expression

Results

Name	Description
`evaluation`	The value based on the applied expression

JSONPath To Parameter

Takes the first value returned by a valid JSONPath expression and sets as an output property

Slug: jsonpath-to-parameter
Type: template
Revision: 3
Image: Default task image.

Parameters

Label	Name	Type	Required	Default	Helper text
Json	`json`	Text Area	No		Json object to query
Query Expression	`query`	Text	No		A valid JSON path expression

Results

Name	Description
`evaluation`	The value based on the applied expression

Example

An earlier task named scan returns a JSON report in its result report, such as {"findings":[{"id":"CVE-2026-1234","severity":"high"}],"total":1}. Pick out the first finding's severity, for a later task to act on.

Field	Value
Task name	`severity`
Json	`$(tasks.scan.results.report)`
Query Expression	`$.findings[0].severity`

The task returns:

Result	Value
`evaluation`	`high`

The task returns the first value the expression matches. An array or object comes back as JSON text, and a number as its digits. For a plain path you don't need this task: a reference can follow a JSON path into a result, such as $(tasks.scan.results.report.total). Use the task for an expression a reference can't take, such as a filter.

Common problems

Symptom	Cause	Fix
The task succeeds, but `evaluation` is empty, and the log shows `TypeError [ERR_INVALID_ARG_TYPE]`	The expression matched nothing.	Check it against the JSON. Expressions start at `$` and are case-sensitive.
The task fails with `Invalid JSON string passed to task`	Json isn't valid JSON, such as an object with unquoted keys.	Check what the earlier task returned, under View Details on the run page.
The task fails with `Invalid mandatory parameters. Check log for details`, after `The parameter 'query' is not defined or empty`	The task needs both Json and Query Expression, although the form marks neither as required.	Fill in both fields.

MongoDB Query Execution

Executes a provided query against a mongo databases endpoint.

Slug: mongodb-query-execution
Type: template
Revision: 1
Image: rtsp/mongosh

Parameters

Label	Name	Type	Required	Default	Helper text
MongoDB database host	`host`	Text	Yes		A valid, reachable mongo db endpoint
MongoDB database port	`port`	Number	Yes
Database name	`database`	Text	No		The database name used in the connection string
Database username	`username`	Text	No		The username to use for running the query
Database password	`password`	Secured	No
TLS enabled	`tlsenabled`	Boolean	No	`false`
Authentication mechanism	`authenticationMechanism`	Select	Yes	`DEFAULT`
Query	`query`	Text Editor - JavaScript/JSON	Yes
Response file path	`ouputfile`	Text	Yes		File path to store the query's output response

No results.

Run Python Script

Execute a Python script. The returned result of the script is returned as the result value of the task.

Slug: run-python-script
Type: script
Revision: 2
Image: boomerangio/worker-python:0.0.3

Warning

Revision 1 does nothing and ends succeeded: the command it calls is no longer in the task image. Later revisions work. Upgrade a node pinned to revision 1 in the editor. See Some older task revisions do nothing.

Parameters

Label	Name	Type	Required	Helper text
Python Version	`pythonVersion`	Select, read-only	Yes	Python 3 or Python 2 only.
Additional Python Packages	`pythonPackages`	Text Editor - Text	No	"Requirement specifier format": www.ibm.biz/BdfMmj
Python Arguments	`pythonArguments`	Text	No	Command line arguments.
Python Script	`pythonScript`	Text Editor	Yes	Python script to execute.
Output file path	`outputFilePath`	Text	No	File path where the script output will be written.

Results

Name	Description
`output`	Python script standard output standard and error content.