Lots of "socket connection timed out on receiving data" errors for Script listener

[alexmears]

Currently

I have script listeners setup on my Jira Cloud that automate updating fields for Jira tickets.

These listeners have been working fine for over a year.

Issue

In the last month or so, we have been getting many errors with the following pattern:

CorrelationId: XXXXXX
RUN Script identifier: YYYYYYYY com.adaptavist.sr.cloud.events.WebhookExecution (jira:issue_updated webhook fired for issue AA-5407) 
Script name: Automated fields Took: 34100ms Logs:
2020-01-31 02:45:39.898 ERROR - Socket connection timed out on receiving data - GET /rest/agile/1.0/issue/AA-5407. The Unirest timeout defaults can be extended as described on this page:  https://github.com/Mashape/unirest-java#timeouts. Please remember though that the maximum invocation time of your code is 30 seconds.
2020-01-31 02:45:39.899 INFO - GET /rest/agile/1.0/issue/AA-5407 asObject Request Duration: 30178ms
2020-01-31 02:45:39.984 ERROR - java.net.SocketTimeoutException: Read timed out
2020-01-31 02:45:40.020 ERROR - Class: com.adaptavist.sr.cloud.events.WebhookExecution, Config: null

 These errors occur, every time we move tickets from one sprint to the next (i.e. on Sprint close), upon which we get around 10+ such errors. They also occur throughout the week at random intervals.

1. What is the possible cause of these errors only recently occurring? We have not changed anything in our scripts recently.

2. How do we fix this?

[Jon Bevan [Adaptavist] · Accepted solution]

Hi Alex,

We're working on a couple of different changes to our infrastructure that should reduce the number of these errors that you see. It is something we are actively working on.

Thanks,

Jon

[alexmears]

@Jon Bevan [Adaptavist] , ok, that's good, but in the meantime, we are having missing and incorrect data in our Jira instance which is causing people to lose trust in the system.

What's our workaround? Should we move our scripts to a more powerful external provider e.g. AWS?