APDEX ETL General Information ¶
General Concept and Environments ¶
Kubed ¶
- JFK - https://apdex.apdex.svc.int.wjfkgen-p-1.prod.infra.webex.com/apdex/home
- JFK Backfill - https://apdex.apdex.svc.int.wjfkgen-p-1.prod.infra.webex.com/apdex-backfill/home
- DFW - https://apdex.apdex.svc.int.wdfwgen-p-2.prod.infra.webex.com/apdex/home
- DFW Backfill - https://apdex.apdex.svc.int.wdfwgen-p-2.prod.infra.webex.com/apdex-backfill/home
ESaaS ¶
- JFK - https://esaas-jfk02-stap-s.webex.com/stap/app/discover
- DFW - https://esaas-dfw02-stap-s.webex.com/stap/app/discover
- Global VIP: https://esaas-stap.webex.com/stap/
General Operations ¶
Deployment ¶
Airflow is (generally) deployed using Jenkins, both for the KubeD environment as well as our Rancher clusters. Details of this are documented in the deployment page.
Failover ¶
The list at the top of this page shows several Airflow APDEX environments. Conceptually we think of our production environment as having a 'primary' and then a secondary or 'failover' environment available for redundancy -- each in a different data center. We refer to the procedure of switching from our 'primary' environment to the secondary of 'failover' environment as 'failing over'. Going back to primary is 'failing back'.
For the Backfill clusters we treat this as the inverse: we tend to keep the DFW cluster running but we also have a JFK Backfill we can run as well as needed.
Backfill Data ¶
Sometimes it is necessarily to re-process some part of our source data to regenerate APDEX scores. We call this 'backfilling'. The process for this is detailed in this page.
General Troubleshooting ¶
Sometimes it will be necessary to check with the team working on Kubed. Refer to the page on seeking help from the WebexKubed Team.