ForgeFX Simulations
Investigation preview
Updated 23 Jul 2026 · 16:28 HST
Azure Monitor / Build Storage

The Alert Is Real.
The Account Is Not Full.

The build account crossed ForgeFX’s static 2 TiB warning line. This is a retention and cost signal, not Azure’s physical capacity ceiling. TeamCity shows one recent local-agent disk-pressure event, but no build failure caused by disk exhaustion.

Current inventory2.005 TiB3,172 blobs across 13 containers
Alert threshold2.000 TiBStatic warning configured by ForgeFX
30-day growth113.1 GiBAverage 3.78 GiB per day
Older than 180 days63.7%1.277 TiB of the current footprint

What Triggered It

Azure Monitor value: 2,203,026,754,557 bytes at 23:11 UTC

Recent build accumulation

Files modified in the last 30 days total 115.1 GiB. John Deere contributes 80.1 GiB, John Deere Pico 28.5 GiB, and John Deere Windows 5.2 GiB.

No automated retention

The account has no lifecycle management policy. Every listed blob remains in the Hot tier. Blob versioning is off; there are no snapshots or retained deleted blobs inflating the inventory.

Age Of Stored Builds

Current blob inventory, grouped by last-modified date

115.1 GiB0–30 days · 5.6%
239.8 GiB31–90 days · 11.7%
390.4 GiB91–180 days · 19.0%
264.0 GiB181–365 days · 12.9%
1.020 TiBOver 365 days · 50.8%

Where The Space Is

The top three containers hold 86.9% of all data

ContainerSizeShareBlobsLatest write
john-deere0.858 TiB42.8%93423 Jul 2026
holotrainer0.562 TiB28.0%97321 Jan 2025
halliburton0.322 TiB16.1%73811 Dec 2025
john-deere-pico0.136 TiB6.8%10923 Jul 2026
john-deere-windows0.080 TiB4.0%6823 Jul 2026

TeamCity Disk-Space Cross-Check

208 finished builds reviewed across the last 30 days

No disk-caused build failure found

Forty-seven builds failed in the review window. Their logs and failure summaries showed tests, execution timeouts, Unity failures, and an upload-version conflict—not “disk full,” “no space left,” or ENOSPC failures.

One agent did run low

John Deere Unit Tests build 715 started on VM-05_BA-03 with 1.54 GiB free on its D: workspace. TeamCity cleaned it to 13.61 GiB before tests ran. The build later failed four tests, so local disk pressure was real but not causal.

EvidenceFindingConclusion
Azure Blob Storage2.005 TiB against an internal 2 TiB alertRetention / cost threshold
TeamCity agent workspace1.54 GiB free, auto-cleaned to 13.61 GiBTemporary local disk pressure
Build failure causeFour unit tests failed after cleanup completedNot caused by disk space

Open TeamCity build 715

Recommended Response

No deletion or configuration change has been made

Approve retention by project

Ask each active project owner how many successful builds must remain immediately downloadable.

Tier before deleting

Move approved older builds to Cool or Archive first. The clearest review pool is the 1.277 TiB older than 180 days.

Automate the policy

Add lifecycle rules after owner sign-off, then replace the misleading “full” description with “crossed the 2 TiB operating threshold.”