Skip to main content

IBM Verify Counts

Daily summary tables in the ibm_verify_events_transformed schema. Use these for IBM Verify reporting.

A nightly Glue job builds them from the raw event stream. It runs at 04:00 ET and writes the previous ET day. Each table is built by one compute_*() function in ibm_verify_events_metrics_daily.py. That code is the definition of every column. The repo is internal, so the link needs cds-snc access.

Data starts on 2026-04-22, the public launch.

Tables

Table Description
ibm_verify_events_transformed.events_auth_total_logins Daily CanadaLogin-wide authentication event totals.
ibm_verify_events_transformed.events_app_login_counts Daily SSO event counts by application.
ibm_verify_events_transformed.events_mfa_activity Daily multi-factor authentication activity by factor and result.
ibm_verify_events_transformed.events_adaptive_risk_summary Daily adaptive access evaluations by risk level.
ibm_verify_events_transformed.events_application_mapping Each application’s IBM Verify ID, current name, and first and last event this year.
No matching items

Dates are strings

from_date and to_date are YYYY-MM-DD strings, not timestamps. Comparing them to a date fails with TYPE_MISMATCH, so cast first:

WHERE CAST(from_date AS DATE) BETWEEN DATE '2026-09-01' AND DATE '2026-09-30'

In R, wrap both sides in as.Date(): filter(as.Date(from_date) >= as.Date("2026-09-01")). A bare string on the right fails the same way, because noctua (like RAthena) turns ISO-date strings into DATE literals.