Enrollment Ops · Incident Digest

July 13 – 16.

Reporting window closes 4:00 PM · SY2026 T1

Total incidents

49

Logged across the support queue

Tech issues

34

69% of all incidents

Operations

15

31% of all incidents

GSA issues — who owns them

Our side (§1 · §3 · §4)

70

52 SFTP · 11 SF-sync · 7 Abanse

MMCL side (§2)

12

No-GSA re-request queue

⚠ Caveat

These logs came from two different reporters. Application IDs repeat across the 14th–16th, so the raw counts likely double-count. Treat as an upper bound until de-duplicated by APP-ID.

↓ scroll

02 · What broke

Five recurring themes

Every ticket in the queue collapses into one of these buckets. The distribution is heaviest at the top.

01

Document upload failures

POP / POE / ID uploaded but not reflecting; missing folders; reattach blocked.

02

Stuck application stages

PPC button missing, frozen at confirmation, admitted, validated, under review.

03

Payment mismatches

Payment processed but status stuck; Maya / CC errors; PPC carry-over.

04

Data-sync integrity

Multiple folders, guardian info not synced, stale apps surfacing in EM.

05

Account & access

Password resets, 404s, endless loading, no PPC button surfaced.

One sentence: technical & operational breakdowns in the enrollment pipeline — clustered on uploads, stage progression, payments, and access.

03 · Root cause analysis

Same symptoms, different roots

Document upload failures

  • Stale EM cache. Post non-Kafka API shift, some browsers hold an old EM build. Fix: hard-refresh. We lowered cache lifetime as a stopgap — may cost some server performance since cache cushions the load.
  • Browser / network + Drive limits. Not new. Google Drive API rate limits + large file sizes. Some tickets get tagged resolved though backend did nothing.

Stuck application stages

  • PPC button. Two roots, same symptom: (1) SF → EM DB data-sync lag, (2) the same stale-cache issue above.
  • Confirmation / admitted freeze. CAMU automation. Our integration is unofficial — we mimic a user browsing CAMU's UI, so when CAMU slows, the automation fails.

Payment issues

  • MMCL portal. Retry after ~1hr usually clears it. Root unclear but sits on MMCL's side.
  • GSA / assessment mismatch. Early-bird applicants missing their 20% Asenso. Because payment already went through, correction runs a long process instead of a simple un-finalization.

Data sync & GSA

  • Multiple folders. Kafka-migration artifact — each file spawned a separate folder. Already fixed.
  • GSA issues. Mostly CAMU — automation fails or gets blocked. Team must manually re-trigger the sync (bandwidth also split toward V2).

Recurring backbone →

STALE CACHE CAMU AUTOMATION SF ↔ EM SYNC LAG DRIVE RATE LIMITS