Skip to content
Data & Database

BlueOrbit Infotech cut database spend by 34% through right-sizing and tiered storage

A 320-person enterprise services firm was running oversized production databases, idle read replicas and provisioned IOPS it did not need. Database spend fell from ₹9.5L to ₹6.3L a month with no loss of performance.

  • Enterprise application services
  • 320 people · Large
  • Hyderabad
  • 6 weeks

Monthly spend

₹9.5L₹6.3L
Saved / month
₹3.2L
Reduction
34%
Payback
13 days

−34%

Database spend

₹38.8L

Annual saving

9

Replicas retired

−22%

p95 query latency

The challenge

BlueOrbit manages managed-database estates for its own platforms and for several enterprise clients. Instances had been sized for peak loads that occurred two days a month, and every database carried the same premium storage regardless of access pattern.

The data engineering team was too busy with client work to revisit sizing decisions made years earlier.

What we reviewed

  • 64 managed database instances and 22 read replicas
  • CPU, memory, connection and IOPS metrics over 90 days
  • Slow-query logs for the 15 costliest databases
  • Backup retention and snapshot policies
  • Non-production database copies

The approach

  1. 1

    Right-size to the real peak

    Instances were resized against 90-day percentiles, with scheduled scaling for the two monthly reporting peaks.

  2. 2

    Remove idle replicas

    Nine read replicas carried almost no traffic; their reads were moved to the primary or a shared replica.

  3. 3

    Fix the expensive queries

    Indexes and query rewrites for eleven heavy queries cut CPU on the busiest clusters by a third.

  4. 4

    Tier storage and backups

    Provisioned IOPS moved to general-purpose storage where safe, and backup retention aligned to contractual requirements.

Before and after

Monthly cost by line item, in Indian rupees.

BlueOrbit Infotech — monthly data & database spend

Line itemBeforeAfterSaving
Production database instances₹4,20,000₹3,08,000−₹1,12,000
Read replicas₹1,60,000₹1,02,000−₹58,000
Provisioned IOPS storage₹1,40,000₹96,000−₹44,000
Backups and snapshots₹95,000₹53,000−₹42,000
Non-production databases₹1,35,000₹68,000−₹67,000
Total₹9,50,000₹6,27,000₹3,23,000 (34%)

Monthly spend over time

₹38.8L saved per year

Two months before the engagement, then six months after the plan was delivered.

Timeline

  1. Weeks 1–2

    Telemetry

    Ninety days of performance data and query logs collected.

  2. Weeks 3–4

    Recommendations

    Per-database plan reviewed with application owners.

  3. Week 5

    Plan delivered

    Change sequence with maintenance windows and rollback steps.

  4. Week 6

    Support

    First wave of resizes executed with our team on call.

“Our databases are faster and a third cheaper. The query fixes alone were worth the engagement, and the resize plan was careful enough that our clients never noticed a change.”
SRSrinivas ReddyHead of Data Engineering, BlueOrbit Infotech

Find out what your agency could save

Book a 20-minute call. We'll show you where the quickest wins usually are for a team like yours, with no data required and no obligation.

  • Free 20-minute assessment
  • NDA before any data is shared
  • Fixed-scope quote, 30/70 payment terms
  • 7 days of post-delivery support
Chat with us