Managed PostgreSQL in Mumbai: pricing, backups and point-in-time recovery

Originally published on the Plexivo blog.

A database is the one part of your stack you cannot afford to lose. So the useful questions about managed PostgreSQL are concrete ones. Where does it run? What does each size cost? How far back can I restore, and how long does that take? Who can connect to it?

This guide answers each of those for Plexivo Postgres, the managed PostgreSQL in Plexivo's Mumbai region. Every number below comes from Plexivo's published price book, documentation or measured drills.

What is Plexivo Postgres?

Plexivo Postgres is managed PostgreSQL. Plexivo looks after the servers, security patches and backups; you look after your data. You choose PostgreSQL 18, which is recommended and chosen for you, or PostgreSQL 17. Updates within a version arrive in your weekly maintenance window, and a move to a newer major version is a guided upgrade you start when you are ready.

Every database includes:

• A database named app, owned by a user named app, with a generated password.
• TLS on every connection. Plexivo Postgres refuses any connection without it.
• Continuous backups and recovery to any second within your backup period, always on.
• The request roles anon, authenticated and service_role, ready for the Data API.

That last point matters if you are building an app from scratch. One Plexivo Postgres can also give you user sign-in, a Data API, Realtime, GraphQL and Functions. See build a backend.

Where does my PostgreSQL database run in India?

Your database runs in Mumbai. Its backups are kept in Mumbai with a second copy in Hyderabad, so a disaster in one city cannot take both. India (Mumbai) is live today, and more countries are available on request. Read the data residency page for where each kind of data lives.

How much does managed PostgreSQL cost in India?

Each size has a monthly price in US dollars, charged by the day. A full day costs the monthly price divided by the days in that month, so a whole month costs exactly the monthly price. Here is the standard family, each with 10 GB of storage:

Size	vCPU	Memory	Price a month
nano	0.25	512 MB	US$11.98
micro	0.5	1 GB	US$22.05
small	1	2 GB	US$42.20
medium	2	4 GB	US$82.50
large	4	8 GB	US$163.10

High-memory and ultra-memory families are there when your working set outgrows the standard ratio. The resource price sits beside your organisation's plan: Developer is US$9 a month billed yearly, Startup US$99 and Business US$599. Indian customers pay in rupees by UPI, card or net banking at this month's rate, plus 18% GST, and receive a GST tax invoice. Every size and limit is on the pricing page.

How do backups work, and how long are they kept?

Backups start the moment a database is created and never stop:

• Continuous archiving. Every change is archived as it happens and reaches the backup within about five minutes.
• Full backups. One every night, and one as soon as a new database starts.
• Kept for your backup period. You choose Backups kept when you create the database, up to your plan's maximum.

Plan	Backups kept, at most	Read replicas per database
Developer	7 days	—
Startup	14 days	1
Business	30 days	2
Enterprise	35 days	Up to 5

Before a large change, choose Back up now on the Backups tab for an extra full backup. Your data is always yours to take, too: Export now prepares a standard PostgreSQL dump that pg_restore reads, and a scheduled export can send one daily or weekly to an S3-compatible bucket you own. The full detail is in backups and restore.

How does point-in-time recovery work?

Point-in-time recovery brings back your data as it was at any moment within your backup period, from the earliest restore point to about five minutes ago. A restore always creates a new database with its own password. Your current database keeps running unchanged, so you can check the copy, then point your app at it or copy back the rows you need.

This is the calm way to undo a mistake. Dropped a table, or ran an update without a filter? Restore to a minute before it into a new database and copy the rows back.

Plexivo measures this rather than describing it. In a drill on 15 October 2026, the restored Plexivo Postgres database was running after 102 s, holding exactly the rows written before the target second. Results like this are published on the reliability page.

What about high availability and failover?

A standby is a second, up-to-date copy of your database. If the primary fails, Plexivo promotes the standby automatically, at the same address and with the same user and password, so your app reconnects without a change. The standby confirms every committed transaction before your app hears that it committed, so a failover loses no committed data.

On Business, every database includes a standby. On Developer, Startup and Enterprise, tick Standby with automatic failover when you create the database; it appears in the price before you create it. In a drill on 14 October 2026, a high-availability failover was writing again in 6.2 s, with all 128 acknowledged writes kept. Read replicas add a read-only address for reports and dashboards. See standby and read replicas.

How many connections can each size take?

Each size has its own connection limit, shown on the create form and on the Settings tab. Plexivo's own monitoring and backups use separate, reserved connections, never yours.

Memory	Connections	Clients through the pooler
512 MB	50	200
1 GB	90	400
2 GB	150	800
4 GB	250	1,500
8 GB	400	3,000

On Business and Enterprise, clients connect through a built-in PgBouncer pooler in transaction mode, at the same host and port. Many client connections then share fewer database connections, and connections are held through a failover instead of dropped. Without the pooler, keep your app's pool below the limit; with several app instances, the limit is shared between them.

Who can connect to my database?

You choose on the database's Network tab, under Access:

• Public, with TLS. Reachable from the internet with its password, over an encrypted connection. New databases start this way.
• Trusted networks. Limit a public database to addresses you list, or choose Add my IP. For a server elsewhere, list its outbound IP addresses.
• Private. No internet address. Apps and Kubernetes workloads in the same project, the Plexivo console, the Data API and the Plexivo CLI tunnel reach it.

The CLI tunnel works with any of these settings, so your laptop never needs to be listed. Details are in network access.

Step by step: create and connect a Plexivo Postgres database

1. Start. In the console, choose Create, then Plexivo Postgres.
2. Name it. Enter the Database name and choose the Project. Apps in the same project reach the database over the private network.
3. Pick a version. PostgreSQL 18 is marked Recommended.
4. Choose size and recovery. Under Size and recovery, pick the Database size, the Storage, Backups kept, and whether you want a Standby with automatic failover. The form shows Connections at this size.
5. Check the price and deploy. The Estimated monthly cost is shown on the right. Choose Deploy.
6. Copy the connection string. On the Connect tab, choose Reveal, then Copy, and Download CA certificate.
7. Set network access. On the Network tab, choose Public, with TLS or Private, add any trusted networks, and choose Save network access.

A revealed connection string looks like this; always use the port shown on the Connect tab:

postgresql://app:<password>@orders-db-3f9a1c.db.plexivo.app:31542/app?sslmode=verify-full&sslrootcert=plexivo-postgres-ca.pem

In Node.js, reuse one pool rather than opening a connection per request:

import pg from 'pg';
const pool = new pg.Pool({
 connectionString: process.env.DATABASE_URL,
 ssl: { ca: process.env.DATABASE_CA_CERT },
 max: 10
});

From your laptop, the Plexivo CLI opens an encrypted tunnel, even to a private database:

plexivo auth login
plexivo db connect orders-db

FAQ

Is there managed PostgreSQL hosted in India?

Yes. Plexivo Postgres runs in Mumbai, and its backups are copied to Hyderabad, both in India. You can choose PostgreSQL 18 or PostgreSQL 17.

How far back can I restore my database?

To any moment within your backup period, up to about five minutes ago. The period is up to 7 days on Developer, 14 on Startup, 30 on Business and 35 on Enterprise. A restore always goes into a new database, so the original keeps serving.

Does a restore overwrite my live database?

No. Every restore creates a new database with its own password. You check the copy, then point your app at it or copy back the rows you need.

Do I need a connection pooler?

On Business and Enterprise, a built-in PgBouncer pooler in transaction mode is on by default. On other plans, keep your app's connection pool below the limit for your size, shown on the create form and the Settings tab.

Can I move my existing PostgreSQL database to Mumbai?

Yes. Open Migrations and choose New migration. The copy reads your source while it keeps running, and you switch your app over when you are ready. See move to Plexivo Postgres.

Create your database

Open cloud.plexivo.io and create a Plexivo Postgres database in Mumbai, with backups from its first second. The free trial is on the Developer plan, and no credit card is required. For a full app, read deploy Next.js with a database in India.