Jitterbit

Root Cause Analysis

We apologize for the service disruption. We appreciate your patience and understanding. We are fully committed to resolving this and minimizing further disruptions.

Issue: The cloud agents scheduler service failed to restart after a scheduled upgrade, disrupting customer schedules during the outage.

Impact: Major
Services Impacted: Customer scheduled operations using Jitterbit Sandbox and Production Cloud Agent groups
Location: APAC Cloud

Problem Description: Jitterbit iPaas scheduler service utilizes an external library for handling timezone-related functions. This library, upon initialization, attempts to update its timezone data file. An incorrect value within this file caused the Scheduler service to fail during startup.

Timeline:
9/6/2024 (UTC)

04:00 - The upgrade process began and the old servers were drain stopped from service

05:15 - The process was finished and internal testing began and discovered the issue with the scheduler service

08:20 - Identified the issue and started to develop a workaround and then apply to the Cloud Agents

08:55 - Workaround applied and confirmed that issues was mitigated

Root Cause: A syntax error was found in third-party timezone data file

Action:
Immediate Action:

• Implemented a work around by fixing the timezone data file and restarting the service
• Document workaround in release notes for customers that may run into this issue with their private agent

Strategic Action:

• A future agent version will not update the timezone data file automatically. 
• Correct syntax of timezone data file