Skip to content

1_ Documentation

Kai Pak edited this page Dec 6, 2022 · 19 revisions

Introduction and Overview

These analytics dashboards primarily use Athena and QuickSight.

Data Pipeline and Backend

SQL Queries in Athena

.
├── atm
├── geo
├── img
├── naif
├── ppi
├── rings
└── sbn

For NoSQL database backend in Athena/Presto, a fairly simple database design is used based on a central raw table that contains all log entries for PDS nodes. The data is stored in S3 by node and this directory structure creates the partition structure for Athena for performance reasons. Aggregated views per node have been created that allow for the identification of requests for tools or specific features of nodes.

Nodes

Engineering (en)

Contains logs relevant to engineering node which is like a central management node for all of PDS. Most of the directories are no longer updated with recent logs as the architecture has changed.

.
├── en-ftp-pds3-report
├── en-ftp-pds4-report
├── en-ftp-pds-proxy
├── en-ftp-report
├── en-http-pdscloud-prod1
├── en-http-pdscloud-prod2
├── en-http-pds-proxy
├── en-http-pdsweb1
├── en-http-pdsweb2
├── en-tomcat-pdsdlb1
├── en-tomcat-pdsdlb2
└── old
directory last_updated
en-ftp-pds3-report August 7, 2019
en-ftp-pds4-report August 7, 2019
en-ftp-pds-proxy September 5, 2019
en-ftp-report March 6, 2017
en-http-pdscloud-prod1 Recent
en-http-pdscloud-prod2 Recent
en-http-pds-proxy April 8, 2022
en-http-pdsweb1 September 2, 2022

Creating Athena Tables

Dashboard and Data Visualization Layer

Troubleshooting