log-tailor

README

log-tailor

A command line program for tailing GCP

Install:

go install github.com/zonkhead/log-tailor@latest

Options:

Usage of ./log-tailor:
  -f value
    	Filter expression (multiple ok)
  -format string
    	Format: json,yaml,csv (default "yaml")
  -l value
    	Log to tail (short name, multiple ok)
  -limit int
    	Number of entries to output. (default 9223372036854775807)
  -p value
    	Project ID (multiple ok)
  -version
    	Show version info

Command line args override values in the config. You can pass a YAML config in via stdin. Here's a simple example for now:

# all or drop-no-match (defaults to all)
match-rule: drop-no-match

projects:
- testing-proj

filters: []
# - protopayload.authenticationInfo.principalEmail = "system:gcp-controller-manager"

common-output:
- timestamp: timestamp
- project: resource.labels.project_id
- log-name:
    src: logName
    regex: projects/.*?/logs/(.*)
    value: $1
- resource-type: resource.type
- labels: resource.labels

logs:
- name: cloudaudit.googleapis.com/activity
  type: k8s_cluster
  output:
  - payload: payload.protopayload
  - principalEmail: payload.protopayload.authenticationInfo.principalEmail
- name: cloudaudit.googleapis.com/data_access
  output:
  - payload: payload.protopayload

CSV output was added to pipe data into SQL databases. If you have tree data being output in your yaml config, it will output it as json so you can use json columns in postgresql, for example. There's an example script that reads from stdin and sends to postgresql:

./log-tailor -format csv < output-config.yaml | ./scripts/csv2psql.sh user_name db_name table_name

Where it is now

You can specify logs, filters, projects, and output formats. If you want to customize (tailor) the output, you can specify a YAML config that maps values from the log entries to keys and values in the output.

Where it may go

Right now it's project-based. There are logs at the billing-account and folder and organization level too. Support for those needs to be added.