Enable Subscription Access for Data Export + Cost Optimization for Cosmos DB

Question

ISSUE SUMMARY

Subscription is disabled, blocking application development. Need temporary re-enablement to export Cosmos DB data and cost optimization guidance.

BUSINESS IMPACT

• Development blocked due to disabled subscription
• Current cost ₹13,600/month for 35 containers
• Need to reduce to ₹2,500-4,000/month for 100+ containers (70-80% reduction)
• Production deployment timeline affected

CURRENT STATE

• Subscription: Visual Studio Enterprise Subscription – MPN
• Cosmos DB: air-cosmos-prod-db (West US 2)
• Containers: 35+
• Throughput Model: Provisioned (per-container)
• Resource Group: air_rg

REQUEST 1: Temporary Subscription Re-enablement

Duration: 2-3 hours

Purpose: Export all containers to JSON for migration

After export, disable again to stop charges

REQUEST 2: Cost Optimization Recommendation

Current: ₹13,600/month for 35 containers

Target: Support 100+ containers at ₹2,500-4,000/month

Questions:

1. Is Serverless Cosmos DB suitable? Container limits? Monthly cost estimate?
2. Is Shared Throughput better? One database with 400 RU/s for 100+ containers? Cost estimate?
3. Which is cheapest: Serverless vs Shared vs PostgreSQL?
4. Best practices for bulk data migration between Cosmos DB accounts?

REQUEST 3: Migration Strategy

• Should we migrate to Serverless, Shared, or PostgreSQL?
• What Azure tools automate container migration?
• Estimated timeline and cost for migration?

Answer 1

Can Azure Support temporarily re-enable my Visual Studio Enterprise Subscription

(ID: <PII Removed>) for 1 day or less ?

PURPOSE: Take a complete backup of my Cosmos DB account (air-cosmos-prod-db)

Once approved, I'll complete the backup within the 24-hour window.

After backup is done, subscription can be disabled again.

Thank you.

Answer 2

Subject: Request for Azure Cosmos DB Support Call – Cost Optimization and Migration Guidance

Hello Microsoft Support Team,

We would like to request a support call with an Azure Cosmos DB specialist to review our current Cosmos DB implementation and help us identify the best approach for optimizing costs.

We are seeking guidance on the following:

Reviewing our current Azure Cosmos DB configuration.

Identifying opportunities to reduce Cosmos DB costs.

Recommending the most suitable storage option for our application (such as Cosmos DB Serverless, Shared Throughput, Azure SQL, PostgreSQL, or other Azure services).

Planning a safe and efficient data migration with minimal downtime.

Sharing best practices for optimizing our database architecture.

Our goal is to reduce our Azure Cosmos DB costs while maintaining application performance, scalability, and reliability.

We kindly request that you arrange a Teams meeting with an Azure Cosmos DB support engineer at your earliest convenience so we can discuss the available options and define the best migration strategy.

We appreciate your support and look forward to hearing from you.

Thank you.

Answer 3

Hello Dinnemidi Ananda Kumar, It sounds like you have three related asks: (1) getting Cosmos DB exports done while the “subscription is disabled”, (2) identifying lower-cost Cosmos DB options for 100+ containers, and (3) planning a migration strategy.

Below is a forum-friendly answer based on the provided docs.

1) “Temporary re-enable subscription” for Cosmos DB export (2–3 hours)

From the provided documentation, the actionable guidance is to open the right kind of Azure support request so you don’t get blocked:

• Ensure you have a paid Azure technical support plan.
• When creating the case, choose Start Again, and select the correct technical service (NOT “Billing” or “Subscription Management”).
• Proceed and request assistance specifically for the technical issue so the case is not closed as a non-qualifying request.

Note: The docs also state that support requests for non-Azure products or Azure technical issues that aren’t handled under the right category may be closed—so the selection matters.

If you can share which Azure portal blade / action is blocked by the “subscription disabled” state (for example, export tooling, Data Explorer, or ARM actions), you’ll help the support engineer target the issue faster.

2) Cosmos DB cost optimization direction (what we can say from the docs)

Cosmos DB costs are driven by:

• Database operations (compute)
• Storage
• Bandwidth/data egress

Compute cost model: Provisioned throughput vs Serverless

Provisioned throughput (RU/s)

• You set RU/s per database/container.
• You’re billed hourly for the maximum provisioned throughput used in that hour.
• You’re charged even if you don’t run workloads because throughput is dedicated to the provisioned resource.

Serverless

• You don’t provision RU/s up front.
• You’re billed based on total Request Units consumed by database operations over the billing period.

So, for cost reduction, the key question is whether your workload is mostly bursty/variable (serverless may help) or mostly steady/predictable (provisioned + right-sizing may help).

How to reduce provisioned-throughput cost (best-practice actions)

Based on the provided “Optimize the costs…” guidance and portal optimization steps:

1. Measure actual RU consumption vs provisioned RU/s

• Look in the Cosmos DB account Metrics for RU usage patterns.

1. Right-size provisioned throughput
   • If containers are consistently underutilized, reduce RU/s to better match demand.
2. Use autoscale when traffic is variable
   • The portal guidance suggests enabling autoscale so RU/s adjusts within a min/max range.
3. Optimize indexing to reduce storage/throughput overhead
   • By default Cosmos DB indexes by default; tuning index policy can reduce storage overhead and avoid unnecessary costs.
4. Optimize storage with TTL / change feed
   • Use TTL to delete data automatically you no longer need.
   • Use change feed for migration to other containers/data stores (no downtime migration concept is mentioned in the docs).

Ways to estimate and forecast costs

• Use the Azure pricing calculator (and log in for your org’s pricing program, if applicable).
• Use Cosmos DB capacity planner / RU estimation guidance (linked from the docs).
• Use Cost Management → Cost analysis for actuals and forecast:
  • Cost analysis supports forecasted cost projection (up to a year) based on recent usage. The docs also provide explicit billing examples showing how provisioned RU/s and serverless RU consumption are charged, which can be used to sanity-check estimates. Reserved capacity / commitment (potential lever)

One of the provided “cost recommendations” mentions:

• Reserved capacity tiers for Cosmos DB can reduce costs if you can commit to RU capacity.

Referance:

• Cosmos DB pricing & billing models: https://azure.microsoft.com/pricing/details/cosmos-db/
• Cosmos DB billing examples / understand your bill: https://learn.microsoft.com/azure/cosmos-db/understand-your-bill#billing-examples
• Plan and manage costs (Cosmos DB): https://learn.microsoft.com/azure/cosmos-db/plan-manage-costs?wt.mc_id=Portal-Microsoft_Azure_Support
• Optimize provisioned throughput cost: https://learn.microsoft.com/azure/cosmos-db/optimize-cost-throughput?wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#data-access-patterns
• Optimize request cost (reads/writes): https://docs.microsoft.com/azure/cosmos-db/optimize-cost-reads-writes
• Optimize storage cost: https://learn.microsoft.com/azure/cosmos-db/optimize-cost-storage?wt.mc_id=knowledgesearch_inproduct_azure-cxp-community-insider#optimize-cost-with-indexing
• Optimize development/testing cost: https://docs.microsoft.com/azure/cosmos-db/optimize-dev-test
• Optimize cost with reserved capacity: https://learn.microsoft.com/azure/cosmos-db/cosmos-db-reserved-capacity
• Optimize multi-region cost: https://docs.microsoft.com/azure/cosmos-db/optimize-cost-regions
• Provisioned throughput vs serverless background (capacity modes): (linked from plan/manage costs page above)
• Download usage details / usage & charges: https://learn.microsoft.com/azure/cost-management-billing/understand/download-azure-daily-usage
• Cost analysis / forecast (Cost Management): https://learn.microsoft.com/azure/cost-management-billing/costs/quick-acm-cost-analysis
• Azure pricing calculator: https://azure.microsoft.com/pricing/calculator/

Answer 4

Hi @Dinnemidi Ananda Kumar

Regarding your migration strategy questions:

Based on the information provided, we can't recommend migrating specifically to Serverless, Shared Throughput, or PostgreSQL based on cost alone. The appropriate target depends on your workload characteristics, including RU consumption, traffic patterns, latency requirements, availability, and application compatibility.

Serverless is generally a good fit for workloads with intermittent or unpredictable traffic because billing is based on consumed request units (RUs) rather than provisioned throughput.

Shared Throughput is often a better option when multiple containers have continuous low-to-moderate activity and can efficiently share a common RU/s pool.

PostgreSQL should only be considered if your application is being redesigned for a relational data model. It is not a direct replacement for Azure Cosmos DB for NoSQL and would typically require application and data model changes.

For migrating data between Azure Cosmos DB accounts, Microsoft supports the following tools:

Azure Cosmos DB Data Migration Tool – suitable for offline or bulk migrations between Cosmos DB accounts or for importing/exporting JSON data.

Azure Data Factory (ADF) – recommended for orchestrating large-scale or automated data movement between Azure Cosmos DB accounts.

Regarding the migration timeline and cost, there isn't a fixed estimate. It depends on factors such as:

Total data volume

Number of containers and documents

Average document size

Available RU/s during migration

Network bandwidth

Whether the migration is offline or requires minimal downtime

If you can share your approximate database size, current RU consumption, and expected traffic pattern (steady or bursty), we can provide more targeted guidance on the most appropriate migration approach.

References:

Azure Cosmos DB throughput models (Serverless vs. Provisioned): https://learn.microsoft.com/azure/cosmos-db/throughput-serverless

Azure Cosmos DB Serverless: https://learn.microsoft.com/azure/cosmos-db/serverless

Azure Cosmos DB Data Migration Tool: https://learn.microsoft.com/azure/cosmos-db/how-to-migrate-desktop-tool

Azure Data Factory connector for Azure Cosmos DB: https://learn.microsoft.com/azure/data-factory/connector-azure-cosmos-db

• Azure Cosmos DB limits and quotas: https://learn.microsoft.com/azure/cosmos-db/concepts-limits