Environment Stability - RCA

Owned by Asha Vade
Mar 21, 2019


  1. ETL job not being executed for 1.14 pre-prod deployment - steps were missed though provided in the deployment tracker
  2. Swarm update due to change in a particular node by Azure
  3. Refactoring of the variable values (500 variables to 60 variables), Communication to the larger group was not done. 
  4. Peer-review on pre-prod was not for validating this change
  5. Possible risk in going to production



Suggested resolution

  1. Can there be a peer review post deployment ?
  2. Big refactoring activities DevOps changes could be done in off peak hours, engineering must be involved to help validate. Communication must be done to all the stakeholders before hand.
  3.  Can a dry run on a new instance happen ?
  4. Can there be a dev ops process kit for every env ?
  5. Can there be a common deployment request tracker ?
  6. Can there be only weekly two times deployment requests for devops
  7. Can Jira be leveraged for deployment ?
  8. How can there be stories that are moving sprint to sprint be minimised ?