PROVIDER - SaaS-native data protection
Druva
Druva delivers data protection as pure SaaS, covering Microsoft 365, Google Workspace, Salesforce, and endpoints from one console with no backup infrastructure to maintain, billed per user or per data volume.
4pts
Detailed Notes
Pros 5 items
Cons 4 items
Service Overview
Druva is a data protection vendor headquartered in California, and within Backup & Migration it sits in the pure-SaaS camp: no local backup server, no repository to size. Customers install one lightweight agent on an endpoint or complete one API authorisation, and indexing, deduplication, storage, and retention all happen in the vendor cloud.
That shape tilts its coverage toward the half of the data estate that lives in someone else's data centre: mail, OneDrive, SharePoint, and Teams in Microsoft 365; Drive and Gmail in Google Workspace; records and attachments in Salesforce; plus documents on employee laptops. What those datasets share is that the SaaS vendor promises availability, not recovery from accidental deletion, which is why third-party backup became standard filler.
The other dimension it emphasises is being API driven. Every backup and restore action can be triggered over REST, and the console is one front end to that API, so backup status and restore requests plug into existing ticketing or automation instead of forcing daily manual checks in a separate interface.
Key Strengths
- No infrastructure to operate: no backup server to patch and no repository to expand, so routine work shifts from maintaining a backup system to reviewing policy and alerts
- SaaS data in one place: Microsoft 365, Google Workspace, and Salesforce share a console, one retention model, and one export workflow instead of three separate tools
- Global deduplication lowers long-term cost: one copy per block across users and versions keeps capacity growth close to linear as retention extends
- API-first automation: status queries, on-demand snapshots, and restore requests all run through the API, fitting existing monitoring and ticketing loops
- Encryption and compliance retention: data is encrypted in transit and at rest, and legal hold plus export auditing produce the evidence most reviews ask for
Product Ecosystem
Platform capabilities
Capabilities split four ways. SaaS backup covers mail, files, records, and metadata in Microsoft 365, Google Workspace, and Salesforce. Endpoint backup covers documents and configuration on laptops and desktops. Cloud workload protection covers virtual machines and databases on AWS and Azure. Governance supplies retention policy, legal hold, sensitive data discovery, and export auditing.
Restore granularity follows the data type: a single email, file, or record for SaaS; a single file or whole device for endpoints; a volume or full machine for cloud workloads. The console records who restored what and when as an exportable audit trail, which carries more weight in offboarding and compliance checks than a successful-job percentage.
Integrations
- Identity and access: enterprise SSO with role-based split across backup, restore, and export permissions
- Communication and ticketing: alerts delivered by email or webhook into Slack, Teams, or an ITSM queue
- Security stack: SIEM ingestion of audit and anomaly events for correlation
- Automation: a REST API for policy management, backup triggers, restore requests, and report export, scriptable for bulk onboarding
- Storage egress: export of protected data to customer-designated object storage or local archive
Typical scenarios
The first group has moved its office suite entirely to the cloud: mail and documents live in Microsoft 365 or Google Workspace with no on-premises equivalent, and the gap to close is rollback after deletion plus handover on offboarding. The second is distributed teams where devices are scattered and IT headcount is thin, making automated endpoint backup and remote wipe more realistic than a file server. The third is compliance-driven work that must search history by user and time and freeze retention for specific objects during litigation.
Limitations
- Data resides in the vendor cloud, so organizations bound by residency rules or industry regulation must verify available regions and certification scope first
- Per-user or per-volume subscriptions mean cost rises with headcount, and long-retained history keeps billing month after month
- Coverage centres on mainstream SaaS suites, so in-house and niche applications require custom scripts against the API rather than ready connectors
- Access from mainland China relies on cross-border links with no first-party local operations team, lengthening troubleshooting and escalation
Use Cases
- Organizations fully on Microsoft 365 or Google Workspace (★★★★★): closes the gap the SaaS vendor leaves on accidental deletion, with policy and export in one console
- Distributed teams and remote work (★★★★☆): endpoint backup plus remote wipe suits scattered devices and limited IT staff, at the cost of an agent on every machine
- Industries needing legal hold and audit trails (★★★★): retention policy, legal hold, and export auditing form a chain of evidence, provided the residency region satisfies the regulator
- Salesforce and similar business systems (★★★☆): record-level restore helps with operator error and bulk import accidents, and depth depends on connector maturity
- Traditional enterprises built around their own data centres (★★☆): local virtualization and database backup is not its strength, which remains the territory of on-premises software
Pricing
| Item | Price | Notes |
|---|---|---|
| SaaS workload subscription | Annual per-user subscription | Billed per protected user with different rates per application; official quote governs |
| Endpoint backup | Annual per-device subscription | Usually bundled with SaaS workloads and priced higher when bought alone |
| Cloud workload protection | Billed per instance or data volume | Varies with cloud estate size, so model it against the existing cloud bill |
| Long-term retention | Capacity-based add-on | Longer retention means more capacity, and cost compounds as history accumulates |
The table covers subscription structure and excludes implementation and migration. Compare on all-in cost per user per month, since that absorbs application count, retention, and endpoint licensing; comparing a single application rate tends to understate long-term cost.
FAQ
Microsoft 365 already exists, why add backup?
Microsoft 365 guarantees service availability, not data history. Retention expiry, accidental deletion, a compromised account wiping data, and gaps in offboarding handover all fall outside recycle bins and version history, and third-party backup covers exactly that span with copies beyond default retention.
Is data in the vendor cloud safe?
It depends on encryption and compliance scope. Druva encrypts data in transit and at rest and publishes certification details, but encryption alone does not satisfy a regulator. Residency region, audit rights, and cross-border transfer clauses need item-by-item review, especially in finance and healthcare.
Does migrating to Druva require downtime?
Usually not. Authorise the connection, run the initial full backup, then cut the old tool's retention over once history has synced, running both in parallel for a period. The real risk is incremental data inside the cutover window, so confirm both sides share the same baseline at the switch.
How does cost compare with a self-hosted backup server?
Short term the subscription costs more; long term you save hardware and labour. Self-hosted cost hides in storage expansion, patching, restore drills, and on-call rotation, all bundled into the subscription. Organizations with a mature backup team and existing storage assets will find self-hosting cheaper at the margin.
Code Examples
Fetch an access token and query backup status
#!/usr/bin/env bash
set -euo pipefail
# Client credentials and API base come from environment variables; verify paths against official API docs
TOKEN=$(curl -s -X POST "${DRUVA_API}/token" \
-H "Content-Type: application/x-www-form-urlencoded" \
--data-urlencode "grant_type=client_credentials" \
--data-urlencode "client_id=${DRUVA_CLIENT_ID}" \
--data-urlencode "client_secret=${DRUVA_CLIENT_SECRET}" | jq -r '.access_token')
# Pull the latest alerts to feed existing monitoring
curl -s "${DRUVA_API}/alerts?limit=20" \
-H "Authorization: Bearer ${TOKEN}" \
| jq -r '.alerts[] | [.severity, .type, .createdAt] | @tsv'
Tokens typically last hours, so never cache them to disk and refresh before expiry in long-running jobs.
Cross-check directory users against protected users
#!/usr/bin/env python3
# Compare directory accounts with protected users and list the unprotected ones
import json
import os
import urllib.request
BASE = os.environ["DRUVA_API"]
TOKEN = os.environ["DRUVA_TOKEN"]
def get(path: str) -> dict:
req = urllib.request.Request(f"{BASE}{path}", headers={"Authorization": f"Bearer {TOKEN}"})
with urllib.request.urlopen(req, timeout=30) as resp:
return json.loads(resp.read())
protected = {
u["email"].lower()
for u in get("/protection/users?limit=1000").get("users", [])
}
# The directory export is a CSV with one email per line; blank and comment lines are skipped
missing = []
with open("directory_users.csv", encoding="utf-8") as fh:
for line in fh:
email = line.strip().lower()
if not email or email.startswith("#"):
continue
if email not in protected:
missing.append(email)
print("unprotected users:", len(missing))
print("\n".join(missing[:50]))
Pagination arguments and field names shift between API versions, so confirm the page limit before bulk fetching or the tail of the user list will be missed.
Community & Resources
- Druva official site — Product lines and positioning
- Druva documentation — Deployment, policy configuration, and restore guides
- Druva developer portal — API reference, authentication, and samples
- Druva Trust Center — Certifications, data residency, and security statements