Making your logs work for you: Drupal escalation and disaster recovery

This is a hands-on session where we will build a robust, enterprise grade, emergency response framework by example and go from inception to completion. We will create a functional demo using simple tools like the Watchdog to create an emergency escalation and planning process.

Failures and errors happen as part of the development lifecycle but what is important is how you respond before, during and after an event. Guard against regressions within your deployments and decrease the time to identify and resolve performance bottlenecks before they become customer or user issues. 

We will take a brief look at emergency management using the Incident command System and how it can be applied to your Drupal sites. 

Using Loggly in conjunction with New Relic we will take a look at how the Drupal 7's good old fashioned Watchdog can trigger events that provide us with stack traces, error logs and critical context when debugging or performing a root cause analysis on a bug or problem. 

We will also define a basic escalation policy by integrating Loggly and Pagerduty to trigger alerts and response protocols.

And what better way to test everything than to break something? Once we have everything set up we get a look of the system in action. We will create an emergency event in order to trigger the incident and escalation response.

Increase visibility for stakeholders to get a better understanding emergency protocols, improve communication, manage expectations and streamline coordination when an event occurs. 

The topics covered will include:

1. Overview
2. Logging in Drupal & PHP
   1. History of logging within Drupal from Drupal 4.6 all the way to Drupal 8
   2. PHP Framework Interop Group (PHP Fig)
      1. Proposing a Standards Recommendation (PSR)
   3. PSR-3 : A common interface for logging libraries
   4. Monolog
      1. Chain of responsibility logging pattern
      2. Core concepts
3. Emergency Management & Planning
   1. Incident planning and policy
   2. Incident command system
   3. Design an incident plan
4. Debugging and monitoring 
   1. Drupal
      1. Watchdog
      2. Composer manager
      3. Monolog module
   2. New Relic
      1. End-to-end transaction stack traces
         1. Application traces
         2. Front-end transactions 
         3. Background transactions (drush, cron)
      2. Track key transactions, hooks and modules
         1. Error logging and aggregation
      3. Availability and Error Monitoring
   3. Loggly
      1. Advanced Log aggregation
      2. Analytics
      3. Monitoring
      4. Custom Dashboards
   4. Pagerduty
      1. Increase visibility 
      2. Escalation policy
      3. Audit trail

By the end of this session it will be quite apparent why an ounce of prevention will be worth a pound of cure. The preparation and foresight is something the developers who write the code, business owners who commissioned the project and the end users who access your site will greatly appreciate.