Skip to content

  • ✅ We have authoritative, named server inventories with memory/disk sizes for the current CDPH CAIR environments

  • ✅ We have clear counts of prod vs non‑prod servers

  • ⚠️ We do nothave explicit Azure VM SKU sizes for each named CAIR server (those were never captured in the Broadgroup thread)

  • ✅ We dohave a defined Azure PoC deployment model (number of VMs + roles)

Below is the best, defensible server list you can use today , grounded strictly in the sourced data.

1. Current CDPH CAIR servers (authoritative list)

Source: CDPH Server information 1.docx provided by Pratap Machavaram [CDPH Serve...ormation 1 | Word]

✅ Production servers

Server Name Environment System Memory Disk (free reported)
PHDCDTMSUTLEP04 PROD Enterprise 128 GB RAM ~3.9 TB
PHDCDTMSUTLEP11 PROD CalREDIE 128 GB RAM ~3.7 TB
PHDCDTMSUTLIP02 PROD CalLIMS 24 GB RAM ~598 GB
PHENTTMSUTLIP08 PROD eHARS 24 GB RAM ~572 GB

[CDPH Serve...ormation 1 | Word]

✅ Total PROD servers: 4

✅ Stage / Test servers

Server Name Environment Memory Licensed CPs
PHDCDMSUTLES04 STAGE 64 GB RAM 38 (22 used)
PHDCDMSUTLES11 STAGE 64 GB RAM 15 (14 used)
PHDCDTMSUTLIS02 STAGE 16 GB RAM 10 (7 used)
PHENTTMSUTLIS08 STAGE 16 GB RAM 10 (2 used)

[CDPH Serve...ormation 1 | Word]

✅ Total STAGE servers: 4

✅ Development servers

Server Name Environment Memory Licensed CPs
PHDCDTMSUTLID04 DEV 16 GB RAM 38 (24 used)
PHDCDMSUTLID11 DEV 16 GB RAM 15 (15 used)

[CDPH Serve...ormation 1 | Word]

✅ Total DEV servers: 2

✅ Grand total (current CDPH CAIR)

Environment Server count
PROD 4
STAGE 4
DEV 2
Total 10 servers

2. Azure PoC deployment (what gets deployed in Azure)

Source: Azure CDPH Demo.onepart and CDPH-Azure-POC-Cloud-SOW-September 2024.docx [Azure CDPH Demo | OneNote], [CDPH-Azure...ember 2024 | Word]

✅ Azure PoC VM roles (not CAIR‑per‑server mapping)

VM Role Prod / Non‑Prod Purpose
Proxy VM Shared Egress HTTPS / SMTP
Rhapsody (RIE) VM PROD PoC Application server
Rhapsody (RIE) VM NON‑PROD PoC Application server
GitLab Runner VM Shared CI/CD & automation

[Azure CDPH Demo | OneNote], [CDPH-Azure...ember 2024 | Word]

✅ Total Azure PoC VMs: 4

⚠️ Important:
These are platform VMs , not 1:1 replacements for the 10 on‑prem CAIR servers.

3. What we do not have (and should not infer)

To stay defensible with Broadgroup / CDPH:

  • ❌ No documented Azure VM SKU per CAIR server (e.g. D8s_v5 vs D16s_v5)

  • ❌ No explicit mapping: PHDCDTMSUTLEP04 → Azure VM X

  • ❌ No committed future-state prod/non‑prod node counts beyond PoC

These were intentionally left flexible and discussed as right‑sizing post‑migration in the Broadgroup thread.

4. How to present this cleanly to Broadgroup / CDPH

Recommended framing (accurate and safe):

  • Current CAIR environment consists of 10 servers across prod, stage, and dev

  • Memory ranges from 16 GB → 128 GB , with production nodes at the upper end

  • Azure PoC deploys 4 platform VMs (proxy, prod RIE, non‑prod RIE, CI/CD)

  • Final Azure VM sizing and server counts will be consolidated and right‑sized , not lifted 1:1

Perfect — thanks for the screenshots. Because you’ve provided the sizing model explicitly , we can now right‑size CDPH cleanly using your own Azure RaaS T‑shirt framework , without guessing VM SKUs or inventing new cost logic.

Below is a defensible, cloud‑native right‑sizing outcome , grounded in:

  • ✅ Your Azure RaaS sizing table (Nov 2025)

  • ✅ Known CDPH message volumes from the Broadgroup data collection

  • ✅ How you actually operate RaaS in Azure (shared platform, not 1:1 servers)

1. Key input: CDPH actual message profile (from discovery)

From the earlier CDPH data you already collected:

  • CalREDIE: ~180k msg/day (in + out)

  • Enterprise: ~190k msg/day

  • eHARS: ~10k msg/day

  • CalLIMS: negligible

  • Other feeds (Parkinson’s, GDSP): low / unknown

  • CAIR overall processing (including responses): ~1.5M–2M msg/day peak observed

✅ Total comfortably under 5 million messages/day
✅ Nowhere near the >10M/day XL threshold

2. Mapping CDPH → your Azure RaaS T‑shirt sizes

Using your exact table definitions :

RaaS Size Message Profile
Small < 500k msg/day
Medium < 5,000,000 msg/day ✅
Large < 10,000,000 msg/day
XL > 10,000,000 msg/day

👉 CDPH squarely fits “Azure RaaS – Medium”

This is true even with:

  • Growth headroom

  • Dual‑engine active/active behavior

  • Response traffic included

3. Right‑sized Azure RaaS deployment for CDPH

✅ Selected size: Azure RaaS – Medium

From your sizing table:

Included:

  • Medium Dev (512 GB ZRS)

  • Medium Test (512 GB ZRS)

  • Medium Prod (512 GB ZRS)

  • One Azure VPN

  • One Azure Firewall

  • Standard platform services (LB, DNS, etc.)

4. Resulting server footprint (cloud‑native)

🔹 Application layer (what Broadgroup actually cares about)

Environment Instance Type Azure Size Memory Storage
Dev Rhapsody Engine Standard_E2bs_v5 16 GiB 512 GB ZRS
Test Rhapsody Engine Standard_E2bs_v5 16 GiB 512 GB ZRS
Prod Rhapsody Engine Standard_E2bs_v5 16 GiB 512 GB ZRS

✅ 3 application instances total
✅ No more 10‑server sprawl
✅ Scale vertically or horizontally later if needed

🔹 Platform / shared services (unchanged by size)

Component Count Notes
Azure Firewall 1 Mandatory per your model
VPN Gateway (VpnGw1) 1 Supports multiple tunnels
Load Balancer 1 Azure‑managed
ZRS Storage Per env Already included
CI/CD Runner Shared Already part of RaaS

5. Cost position (from your Medium table)

From your Medium RaaS AWS Cost section (Azure‑equivalent):

  • Monthly total: $1,432

  • Annual total: $17,184

  • Excludes margin & labor (as stated in your table)

✅ This aligns exactly with the summary row in your screenshot.

6. Why Medium (and not Large or XL)

Why not Large?

  • Large is sized for < 10M msg/day

  • CDPH is operating at ~20–25% of that ceiling

  • Would be over‑allocating CPU, RAM, and cost

Why not Small?

  • Combined traffic + future growth would push Small

  • Medium gives safe headroom without waste

Why not XL?

  • XL assumes national‑scale throughput

  • CDPH is nowhere near sustained >10M/day

7. Executive‑ready summary (you can reuse this verbatim)

Based on CDPH’s observed message volumes and Rhapsody’s Azure RaaS sizing model, the environment is best right‑sized as Azure RaaS – Medium.

This provides three Rhapsody engine instances (Dev, Test, Prod), each with 16 GiB RAM and 512 GB ZRS storage, and supports up to 5 million messages per day with growth headroom.

This replaces the legacy 10‑server on‑prem footprint with a consolidated, cloud‑native architecture while reducing operational complexity and cost.