Netriva Network Monitoring & Operations
Menu

Installation & device guide

Install Netriva and connect your first device.

A practical walkthrough for customer administrators and ITES deployment engineers. Choose your hosting, prepare the installation, connect a collector and confirm that measurements and alerts are working.

Before you begin

Your administrator or ITES provides your workspace invitation. Accept it, set your password, complete any required MFA and choose the workspace you will manage. Installing Netriva on a server also needs an authorised host administrator.

Have your site names, device management addresses, approved read-only access, alert contacts and backup owner ready. Start with one site and one device before expanding.

Arrange installation support with ITES →

1. Choose who hosts your workspace.

ITES hosting

ITES runs the central application and its monitoring services. You need workspace access and a collector that can reach the equipment at your site.

  1. Agree sites, access, alert contacts and retention with ITES.
  2. Accept your administrator invitation and open your workspace.
  3. Arrange the local collector, then follow the device connection steps below.
Continue to device connections →

Your own server

Run the application and monitoring services on a dedicated Linux server or VM. Agree self-hosted availability and the installation package with ITES before preparing the deployment.

  1. Prepare the host, network access and backup destination.
  2. Receive the package and settings for your installation.
  3. Install and verify the platform, then connect your collector and devices.
Review server requirements →

Customer administrator

Confirm sites, team access, device permissions, alert recipients and retention. For your own server, assign responsibility for host updates, DNS/TLS, backups and recovery.

ITES deployment engineer

Confirm package compatibility, complete installation checks, enrol collectors and prove collection and alert delivery. Customer workspace access uses the permissions and support access agreed with the customer.

2. Prepare a server for the full installation.

This is the current resource baseline for a full self-hosted installation. ITES reviews your device measurements, collection intervals, history retention, reporting and optional features before confirming capacity.

Processor
At least 4 CPU cores / vCPUs
Memory
At least 16 GiB RAM
Storage
At least 100 GiB free persistent storage
Operating system
64-bit Linux on x86_64 / amd64

The baseline is an installation requirement, not a guaranteed device count. A 2-core / 8 GB / 50 GB server needs a separately reviewed deployment profile. A collector-only host is sized for its site; it does not need to host the whole central platform.

Engineer detail: check the proposed host

Run these read-only checks on the proposed Linux host. Check free space on the filesystem selected for Netriva data, not just the root filesystem.

uname -m
nproc
free -h
df -h /srv/netriva
docker --version
docker compose version

The example data directory must already exist. Review the host facts with ITES before installation. Docker's Ubuntu installation guide explains host prerequisites; use the versions approved for your package.

3. Install on your own server.

Work with your host administrator and assigned deployment engineer. The package includes the commands and settings for your installation. Follow this sequence, recording the result of each step.

3.1 · Receive and check your installation package

Obtain the approved package, installation instructions, trusted verification material and any required entitlement from ITES through the agreed delivery channel. Confirm the host architecture and release version match your installation.

Keep the package in a dedicated working directory. Your engineer verifies its signature and file inventory before using the included application and container images. An existing workspace account alone does not install the server.

Ready to continue: the package is verified and the installation instructions match your host.

3.2 · Prepare the protected settings and data directories

Agree the installation root, data volumes, workspace hostname and backup location. Your engineer prepares the environment settings from the package template: application keys, separate database roles, HTTPS material, mail service and collector/telemetry endpoints.

Keep configuration files readable only by the authorised service or host administrator. Enter secrets locally in the protected configuration; keep passwords, private keys and enrolment tokens out of email, chat and shell history.

Ready to continue: protected settings exist and data/backup locations have sufficient free space.

3.3 · Run the host preflight and review the installation plan

From the verified package working directory, your engineer checks the host, protected environment file and combined service configuration. The following paths are examples; substitute the approved paths from your installation instructions.

python3 scripts/self_hosted/manage.py preflight \
  --compose infra/staging/compose.yaml \
  --compose infra/staging/compose.telemetry.yaml \
  --compose infra/self-hosted/compose.yaml \
  --env-file /etc/netriva/self-hosted.env \
  --install-root /srv/netriva

python3 scripts/self_hosted/manage.py plan \
  --operation install --install-root /srv/netriva

The environment file and installation root must already exist. Preflight reports PASS or FAIL. The plan command only describes the work; it does not install or start services. Resolve any failed checks before proceeding.

Ready to continue: preflight passes and your host administrator has reviewed the installation plan.

3.4 · Install the verified release and start the services
  1. Your engineer uses the package's release installer to verify compatibility and create the versioned release with its protected shared settings and storage.
  2. Load the package's approved container images and apply the supplied service configuration. Use the separate database migration role for the installation's migrations.
  3. Start the application, database, queue worker and scheduler, then the collector-control, telemetry storage, rule evaluation and notification services needed for monitoring.
  4. Complete health and access-isolation checks before opening the workspace to users. Use the package's activation and rollback instructions for the installed release.

Ready to continue: the platform and monitoring services are running; a working sign-in page alone does not confirm the monitoring path.

3.5 · Confirm HTTPS, administrator access and backups

Open your network using its agreed HTTPS address. Sign in with the administrator account provided during installation, set your own password and complete the required account protection. Confirm you are in the correct workspace.

Check the mail service, backup schedule, off-server backup copy and recovery ownership. Verify a restore into a separate test environment before relying on the backup. Record the installed version and the update/rollback contact.

Ready to continue: administrator access works, backups are checked and someone owns recovery.

4. Connect your first device.

These steps apply to ITES-hosted and customer-hosted workspaces. Use the links after signing in; each action requires the role and site access assigned to you.

4.1 · Create the site and its network

Open Sites, add the office, branch or data centre and confirm its timezone. Add or select the site's network / VRF, such as Office LAN. A VRF separates routing contexts where the same IP address may be used in different networks.

Choose the correct site before adding a device or collector. Give the engineers who will configure monitoring access to this site.

Open Sites →
4.2 · Install and enrol a local collector
  1. Choose an approved Linux collector host that can reach the device management addresses and the supplied Netriva control/ingestion endpoints.
  2. Your engineer installs the matching ITES collector package, its approved exporters and the service configuration. Agree service restart behaviour and storage for buffering.
  3. Open Collectors, filter to the intended site and select Create one-time invitation. Complete the requested account verification.
  4. Save the invitation locally in a protected file on the collector host. Use the verified control URL and bootstrap CA supplied for your installation.
  5. Enrol the collector, then run it through the installed service. Return to Collectors and confirm an active identity and a current heartbeat.
Engineer detail: collector enrolment command

After the matching collector binary is installed, set NETRIVA_CONTROL_URL to the supplied HTTPS collector-control base URL. The example CA, invitation and state paths must match your installation and be accessible only to the collector service account.

ites-collector enroll \
  --control-url "$NETRIVA_CONTROL_URL" \
  --bootstrap-ca /etc/netriva/collector-ca.pem \
  --invitation-file /etc/netriva/collector-invitation \
  --state-dir /var/lib/ites-collector \
  --name "Office collector"

ites-collector health --state-dir /var/lib/ites-collector

The health command reports local state. Confirm the server has received a current heartbeat as well. Start the installed service with the same state directory; a successfully issued certificate alone does not show that collection is running.

Open Collectors →
4.3 · Prepare read-only device access

For network devices, your administrator can configure a read-only SNMPv3 account with authentication and encryption (authPriv) and allow access from your collector. Use the SNMP version required by the selected profile; restricted SNMPv2c is an option for approved legacy interface monitoring. Follow the manufacturer's instructions for the exact model and firmware.

For compatible Linux hosts or Windows servers, have an authorised administrator install the exporter required by the selected profile. For supported virtualisation APIs, prepare the profile's least-privilege account and verified TLS settings. These operations require the customer's approval.

Save required read-only credentials in Credential vault and select the saved reference when setting up monitoring. Do not put device passwords in the inventory description.

Open Credential vault →
4.4 · Add the device and choose its monitoring profile
  1. Open Devices → Add device. Enter a recognisable name, select the site and network / VRF and save its management IP address.
  2. Select a monitoring profile that matches the information you need: for example, Network device — ports & traffic (SNMP) or Linux server — system resources.
  3. Read the profile guide for its access requirements and supported measurements. Select the saved credential where required.
  4. Set the management state to Managed when the device is ready for monitoring setup, then save.

Saving the device creates inventory. Measurements start after an approved collection target and working collector are configured.

Add a device →
4.5 · Approve the collection target and publish its configuration
  1. Open the collector assigned to this site and choose Approve collection target.
  2. Select the managed device, its saved management address, matching profile and protocol. Confirm the required credential or exporter/API deployment consent.
  3. Set the collection interval and timeout for the approved workload. An interval of 60 seconds means once per minute; it is a starting example, not a requirement for every profile.
  4. Approve the target, then select Publish signed collection revision on the collector page.
  5. Wait for the desired revision to appear as applied. If it is rejected or still pending, review the reported reason before adding more targets.

Ready to continue: the approved target is included in an applied configuration and the collector heartbeat is current.

Open the site collector →
4.6 · Check incoming measurements and configure alerts

Open the device and review its available interfaces, measurements and observation timestamps. Confirm the selected profile's expected data is arriving recently. Missing or stale values require investigation; they do not mean the device is healthy.

Your engineer confirms the rule evaluator and essential notification route. In Response policy, configure collaboration recipients or approved webhooks, severity, quiet hours, maintenance windows and escalation delays. Agree a controlled test and confirm receipt with the intended recipient before handing over monitoring.

Open Response policy →

5. Verify the installation and hand it over.

If a step does not work

You cannot see an action
Check your workspace, role and site access with your administrator. Host administrator access and Netriva workspace permissions are separate.
Preflight fails
Read the failed check. Confirm CPU, RAM, free space, the protected settings file and the complete package configuration. Resolve the cause before installation.
Collector has no current heartbeat
Check the running service, control URL, time, CA/certificate trust and approved network path. An expired or consumed invitation requires the authorised enrolment workflow.
Configuration is pending or rejected
Check the desired/applied revision and rejection reason. Confirm the target uses the correct site, profile, address, protocol and saved credential.
A device has no measurements
Check Managed state, approved target and applied revision, then the collector-to-device path, SNMP permissions or exporter/API endpoint. Do not broaden network access just to make a check pass.
An alert does not arrive
Confirm incoming evidence and the rule evaluator first. Then inspect recipients, severity, maintenance/quiet-hour decisions, notification attempts and mail/webhook delivery status.

Linux device monitoring

Connect CPU, memory and disk measurements.

These steps apply to compatible Linux servers, virtual machines, workstations and appliances that permit node_exporter installation and expose the required metrics. Install the exporter on each Linux device you want to monitor; the Netriva collector reads its approved endpoint.

This section concerns monitored devices. Hosting Netriva itself has separate server requirements. For Linux-based routers, switches or firewalls that do not permit an exporter, use a compatible SNMP or API profile.

  1. 1. Prepare the Linux device

    Your administrator or ITES installs node_exporter 1.12.1 on the monitored device with your approval. Confirm that its build matches the device’s architecture and operating system and exposes the metrics required by the profile.

  2. 2. Approve collector access

    Allow only the intended collector to reach the device’s approved /metrics endpoint on port 9100. Run the exporter with limited permissions and access to the host resources being monitored.

  3. 3. Verify real measurements

    Select the Linux resource profile (labelled Linux server — system resources in the portal), approve collection and check for fresh CPU, memory, filesystem and disk measurements from the intended device. Missing measurements need investigation; they are not healthy or zero values.

Discuss assisted Linux setup with ITES →