Firewall Log Retention in India Starts at 180 Days

  • Home
  • Firewall Log Retention in India Starts at 180 Days
Firewall Log Retention in India Starts at 180 Days

The auditor’s email usually lands on a Thursday afternoon. “Please share firewall logs for the head office and the Nagpur and Coimbatore branches, 3 to 9 May.” You open FortiAnalyzer, scroll back, and the oldest record is from late July. Nobody deleted anything. The disk filled up and the oldest logs rolled off, exactly as configured.

That gap is the finding we see most often when a company asks us to review its firewall log retention in India before audit season. The policy document says 180 days. The appliance quietly keeps whatever fits. The smaller branch firewalls, the ones sitting on a single broadband line, often never sent a log anywhere central at all.

So instead of a theory piece, here are the questions an auditor, a regulator or your own board will actually put to you, and the answers that hold up.

The short answer: under the CERT-In directions of April 2022, you keep logs from all ICT systems, firewalls included, for a rolling 180 days and hand them over to CERT-In on request. Logs of personal data processing need at least one year once the DPDP Rules’ main obligations apply. Sector regulators can ask for more, so treat 180 days as the floor.

Which rule sets the clock, and which one wins?

The CERT-In directions issued on 28 April 2022 apply to service providers, intermediaries, data centres, body corporates and government organisations. They require logs of all ICT systems to be kept securely for a rolling 180 days within Indian jurisdiction. CERT-In’s own FAQs later softened the location part: logs may sit outside India as long as you can produce them to CERT-In in a reasonable time. We still keep client logs in India; it removes an argument you don’t need mid-incident.

The Digital Personal Data Protection Rules, 2025, notified by MeitY in November 2025, add a one-year minimum for logs connected to the processing of personal data. Most business obligations under those Rules, including the security safeguards, take effect 18 months after notification. Firewall logs carry source IPs, usernames from FSSO or RADIUS, and URLs people visited. We treat them as in scope. Your legal counsel should confirm that for your organisation; this is an operational reading, not legal advice.

Then there’s your sector regulator. Banks and NBFCs answer to the RBI, brokers and AMCs to SEBI’s cyber framework, insurers to IRDAI. Each has its own logging expectations, and they change through circulars. Read the current one. If you’re regulated by the RBI, our note on RBI IT governance requirements for banks and NBFCs covers where logging sits in that framework.

The rule of thumb is simple: the longest applicable period wins, and you add a buffer on top for investigations that start late.

Source Who it applies to Minimum period Practical note
CERT-In directions (April 2022) Service providers, intermediaries, data centres, body corporates, government Rolling 180 days Must be producible to CERT-In on request
DPDP Rules, 2025 Data fiduciaries and their processors At least one year for logs of personal data processing Main obligations apply 18 months after notification
RBI, SEBI, IRDAI frameworks Regulated entities As set in your current circular Check before every audit cycle
Your internal policy You Longest of the above, plus a buffer Write down who approved it

Can you produce any single day, from every site?

This is the question that fails people. A retention policy is a promise about completeness, and completeness breaks at the edges of the network.

Picture a 400-seat auto-components company with a plant in Pune, a head office in Gurugram and nine sales branches. The head office FortiGate sends everything to FortiAnalyzer. The plant firewall was added two years later by a different vendor and logs to its own disk. Four branch units are small desktop models logging to memory, which means a reboot during a power cut wipes the history. On paper, retention is 180 days. In practice, it’s 180 days for one site.

Before any auditor asks, walk through this list:

  • Every firewall, including branch and plant units, forwards logs to a central collector rather than keeping them only on local disk or memory.
  • Both members of each HA pair are accounted for, so a failover doesn’t create a hole.
  • Logs from every VDOM reach the collector, not just the root VDOM.
  • SSL VPN and IPsec events, admin logins and configuration changes are kept alongside traffic logs.
  • Cloud firewalls and the stray non-Fortinet box at an acquired site are covered too.
  • Once a month, someone pulls a random date older than 150 days and checks it comes back whole.

That last item is the one almost nobody does, and it’s the only one that proves the others work.

Will your timestamps stand up in an investigation?

Logs with the wrong time send an investigation down the wrong path. The CERT-In directions ask organisations to synchronise ICT system clocks to the NTP servers of the National Informatics Centre or the National Physical Laboratory, or to servers traceable to them. The FAQs clarified that you don’t have to set every device to IST. What matters is that everything agrees.

Here’s where it usually goes wrong. Branch firewalls drift because they were never pointed at a proper NTP source. The collector stores logs in UTC while the Windows servers log in IST. Someone correlates a VPN login with a file server event and is off by five and a half hours, and the whole timeline is wrong.

The fix is boring and takes an afternoon: one approved time source, one documented timezone convention, and a monitoring check that alerts when any device drifts.

Which firewall logs are worth keeping, and at what detail?

Many teams log only denied traffic because it’s smaller. We think that’s the wrong trade. Denies tell you what the firewall stopped. An investigation needs to know what got through: the outbound session to an unfamiliar host at 2 a.m., the steady trickle of data to a file-sharing site, the admin who connected from an odd location.

Our default on FortiGate is to log all sessions on internet-facing policies and on policies between security zones, plus security events from IPS, web filter, application control and DNS filtering. Those are the records that matter when you’re tracing the attacker behaviours we covered in our post on living-off-the-land techniques, where the malicious activity looks like ordinary admin traffic.

What do we turn down? High-volume, low-value internal allow rules: backup jobs between two servers in the same rack, or a monitoring poller hitting every switch every minute. Logging those in full can double storage without adding investigative value. Document the exception, though. An auditor will accept a reasoned decision far more readily than an unexplained gap.

Sizing firewall log retention in India without guesswork

Don’t size from a datasheet. Measure. Take a normal working week, not the Diwali week or a quarter-end close, and note how much log data your collector actually receives per day. Multiply by the retention period, apply the compression your collector or SIEM really achieves, and add room for growth and for the verbose logging someone forgets to switch off.

Then split the storage into tiers. A full year on fast storage is rarely necessary.

Tier Age of logs Used for Typical home
Hot 0 to 30 days Daily SOC triage, quick searches FortiAnalyzer or SIEM fast storage
Cold 31 to 180 days Investigations, CERT-In requests Compressed, cheaper disk, slower queries
Archive 181 days to a year or more DPDP and sector obligations, legal hold Write-once object storage in an Indian region

One honest warning: the archive tier is only useful if you’ve tested getting data back out of it. We’ve seen archives that technically held a year of logs, in a format nobody could search without a week of scripting.

Who can delete the logs, and would you know?

Attackers who reach a firewall admin account often turn logging down or clear it. If the person who manages the firewalls also has full rights on the log collector, one stolen password removes your evidence.

Separate the roles. Give the collector its own admin accounts with MFA, and no shared “admin” login. Send a copy of the logs to a second destination the firewall team can’t modify. Raise an alert whenever a FortiGate’s log settings change, because that config event is a strong signal on its own. Our FortiGate hardening checklist covers the admin-access side of this in more detail.

How PrahiX Ora handles retention and what comes after it

Retention by itself answers the auditor. It doesn’t help your team find anything in six months of firewall records when an incident starts at midnight. That second problem is why we deploy and operate PrahiX Ora for clients. It’s a unified SecOps platform built by PrahiX Tech Pvt Ltd, and PJ Networks is its primary field deployment and operations partner.

The SIEM is where retention lives. It ingests logs and events from firewalls, servers, identity systems and cloud sources, and stores them in hot, cold and archive tiers, so keeping 180 days in-country under the CERT-In direction doesn’t mean paying fast-storage prices for all of it. Correlation rules are mapped to MITRE ATT&CK. Instead of a pile of separate alerts, the platform rebuilds a graph-based storyline of an attack: which host, which account, which outbound connection, in what order.

The NMS side gives the NOC one view of firewalls, switches, access points and WAN or SD-WAN links. It discovers topology through LLDP and CDP and traces network paths. It uses ML-based anomaly detection and auto-healing policies, which matters most on multi-vendor estates where each box has its own console.

Video surveillance (VMS) brings ONVIF, Hikvision and Dahua cameras into the same operations view, with video analytics. For a plant or a retail chain, that puts physical and network events on one screen.

SOAR runs playbooks through pre-built connectors and can push a blocklist to FortiGate automatically. When CERT-In’s six-hour reporting window is running, that automation leaves the analysts time to write the report.

The platform supports compliance with CERT-In and DPDP obligations and helps you evidence them. It doesn’t make you compliant by itself. If you’d like to see it against your own logs, our managed SIEM service page explains how a pilot works.

Frequently asked questions

Does CERT-In require firewall logs to be stored in India?

The 2022 directions say logs should be maintained within Indian jurisdiction. CERT-In’s later FAQs allowed storage abroad, provided the organisation can produce the logs to CERT-In within a reasonable time when asked. In practice, an Indian region is the simpler choice, because it avoids cross-border access questions in the middle of an incident or an audit.

Is FortiAnalyzer enough for 180-day firewall log retention?

It can be, if the appliance or VM has enough disk for your real daily volume and every firewall actually forwards to it. The common failure is sizing for today’s traffic and then adding sites. Check the oldest log date on the device each month, and keep a second copy somewhere else so one failure doesn’t take out your only record.

Do small branch offices need the same log retention?

Yes. The directions talk about all ICT systems, and attackers like small branches precisely because nobody watches them. Small desktop firewalls usually lack a disk, so they need to forward logs to the central collector, ideally over the existing VPN tunnel. Don’t rely on local memory logging, which is lost on reboot.

How long should a bank or NBFC keep firewall logs?

Start with the 180-day CERT-In floor, then read your current RBI master direction and any circulars that apply to your category of entity, since these set their own expectations and can be stricter. Logs that touch customer personal data also fall under the DPDP one-year minimum once it applies. Use the longest period, and confirm it with your compliance team.

One test to run before Friday

Pick a date about 170 days ago. Ask your team, or your provider, to produce firewall logs for that date from three locations: the head office, your largest plant or warehouse, and your smallest branch. Time how long it takes and note what’s missing. That single exercise tells you more about your real retention than any policy document.

If the results are uncomfortable, that’s normal, and it’s fixable. Our team runs firewall audits and rule reviews that include log coverage, retention and time sync. We’re happy to talk through what you find, whether or not you work with us.

Leave a Reply

Your email address will not be published. Required fields are marked *