ScriptRunner: Keep request alive during a script that takes a long time to finish

[Ricardo Martinez]

I have a script that makes a rest api call to another application.
The script then waits for the call to finish.

Meaning the script executes the call then sleeps for 5 seconds then asks if it finished and keeps waiting 5 seconds until it does.

Sometimes the whole execution of the script can take more than 1 hour.

Jira is under a NGINX and this proxy closes the connection after a time out. When this happens the script keeps executing but the console finishes and doesn't come up with results.

Is there a way for the script to keep the connection open while waiting?

Oversimplified example:

while(OtherRestCallIsNotFinished)

{ sleep(1000) keepConnectionOpenWithProxy() }

[Ilya Turov · Accepted solution]

What about storing the results into a separate file? 

Or simply logging them and retrieving this way.

[Ricardo Martinez]

I already do that, but I do not know when the whole script is finished, so I can go to the file and check it.

It would be much better to keep the connection open until it does.

[Ilya Turov]

Guess you can add a piece of code that will notify you via email after it's done :)

import com.atlassian.jira.component.ComponentAccessor
import com.atlassian.mail.Email
import com.atlassian.mail.queue.SingleMailQueueItem

def email = new Email("someuser@example.com")
email.setSubject("all done")
email.setBody("")
email.setMimeType("text/html") // for messages with an html body

SingleMailQueueItem item = new SingleMailQueueItem(email)
ComponentAccessor.mailQueue.addItem(item)

[Ricardo Martinez]

The email alternative is a good idea, I have actually just implemented it.

A way to know when the script finished in case the console closed the connection.

I got a response from Adaptavist (ScriptRunner) about this issue.

https://productsupport.adaptavist.com/browse/SRPLAT-3

Basically it is an open request to have a progress response on the script.

In the meantime an email sounds like a good idea to get notified when it finishes.