Weekly Digest #115

Articles

Logging practices I follow

• Is this log really needed? does it rely important information I couldn’t get from the other logs in the same flow?
• Am I going to log an object that can be huge on production? If so, can I just log a few metrics of that objects instead? for example, it’s length, or handpick a few important attribute to log.
• Does the information I am about to log will help me to debug/understand the flow?

Log levels:

• ERROR: Parts of the flow failed, we want to send alerts to our on-call for this failures.
• WARNING: Doesn’t necessarily point to a failure, but an unexpected behavior that should be investigated.
• INFO: Record major events in the flow to help the developer reading it understand what was being executed.
• DEBUG: Like INFO but more detailed, including inspection into objects, data structures, etc.

Power Struggles Among Nice People

the most successful senior leaders display 1) a relatively high need for power, 2) a relatively low need for affiliation, and 3) a high degree of impulse control, a distinctive pattern that he called “leadership motive syndrome.”