Database › DBaaS

Managed databases that stay fast, backed up, and in India

Spin up MySQL, PostgreSQL or MariaDB in minutes. E2E handles provisioning, patching, failover and replication — you get an SSL-secured endpoint, snapshots, read replicas and a parameter group to tune, billed in INR.

3 engines
MySQL · PostgreSQL · MariaDB
₹9/hr*
Managed nodes from
Delhi ↔ Chennai
Cross-region read replicas
99.9%
Uptime SLA

*Indicative — see the live rate card for current pricing.

The boring database work, automated

Provisioning, patching and 3am failovers are our problem now.

Replication & failover

Every engine ships with replication and in-cluster failover support. For regional failure, one-action promotion turns a read replica into an independent primary and keeps you serving.

Snapshots & backups

On-demand snapshots, API-schedulable snapshots, and restore-from-snapshot on every engine — so "did we back that up?" always has a good answer.

Read replicas

Offload reads to replicas with their own endpoints. Same-region replicas work out of the box; cross-region Delhi ↔ Chennai replicas run over VPC peering. Promotion to master is one action.

Parameter groups

Tune engine configuration — innodb_buffer_pool_size, shared_buffers, max_connections and friends — in reusable parameter groups you attach, edit and sync to running databases.

Locked-down access

SSL-encrypted connections and an allowed-hosts firewall so only the IPs you specify can reach the database. Private networking via VPC attachment.

Monitoring & logs

Cluster monitoring graphs, database logs and lifecycle controls — start, stop, restart, plan upgrades — all from the MyAccount console or the REST API.

Supported engines & versions

As published in the E2E DBaaS documentation.

CapabilityMySQLPostgreSQLMariaDB
Documented versions5.6 · 5.7 · 8.010 · 11 · 12 · 13 · 1410.4
Recommended for new deployments8.01410.4
Fully managed (provision, patch, maintain)
Automatic failover
Replication & read replicas
Read replica promotion
Snapshot & backup support
Parameter groups
Default port330654323306

Plans combine CPU, memory and storage to fit your workload; a small dev instance can start around ₹9/hr*.

How it works

From console to connection string in four steps.

Step 1

Pick engine & plan

Choose MySQL, PostgreSQL or MariaDB and a node plan — the CPU/memory/storage mix that matches your workload.

Step 2

Set access rules

Create the database user, attach a VPC for private connectivity, and allowlist the hosts that may connect.

Step 3

Connect over SSL

Grab host, port, username and database from Connection Details and connect with any standard client — psql or mysql, phpMyAdmin included.

Step 4

Operate & scale

Schedule snapshots, attach a parameter group, add read replicas as reads grow, and upgrade the plan when it's time to scale up.

Data residency isn't a feature — it's the default

Your database never leaves Indian jurisdiction, which makes several hard conversations easy.

DPDP-friendly by design

Data sits in Delhi NCR or Chennai, operated by an NSE & BSE-listed Indian company under Indian law — a clean answer for DPDP Act governance reviews.

Certified infrastructure

SOC 2 and ISO-certified, MeitY-empanelled platform with a public status page.

Next to your AI stack

Keep application data on the same low-latency network as your GPU nodes, Kubernetes clusters and TIR endpoints — no cross-cloud egress, no forex.

Frequently Asked Questions

Everything you need to know about Database as a Service.

Database as a Service

MySQL 5.6, 5.7 and 8.0; PostgreSQL 10 through 14; and MariaDB 10.4. New deployments should use MySQL 8.0 and PostgreSQL 14 — earlier versions are maintained for legacy migrations and have reached upstream end-of-life. Engine configuration is tunable through parameter groups.

Same-region replicas need no special setup. For cross-region (currently Delhi ↔ Chennai), attach the primary to a VPC with an active peering connection to a VPC in the target region. Replicas inherit the primary's size, get their own endpoint, and auto-resync after connectivity interruptions. Note: cross-region replicas aren't available for encrypted instances.

Engines run with automatic failover and replication support. If a region becomes unavailable, promote a read replica — once promoted it becomes the primary of a new independent cluster and serves reads and writes. Promotion is one-way; there's no backward sync to the old cluster.

Yes — connections are SSL-encrypted, and the allowed-hosts firewall limits access to the client IPs you specify. Attach a VPC to keep traffic on private networking between your app nodes and the database.

Upgrade the cluster to a bigger plan for vertical scaling (CPU/RAM/storage), and add read replicas to spread read traffic horizontally. Both are console actions with API equivalents.

Start Building Today

Ship the app. We'll babysit the database.

Launch MySQL, PostgreSQL or MariaDB in minutes.