Skip to content

Configuration Overview

There are three ways to send logs to SenseOn. The correct way for you depends on your environment and your needs. These ways are the same as the three deployment models: Receive, Relay and Retrieve.

Your ingestion endpoint

SenseOn gives you one ingestion hostname for your organisation:

<code-name>-logs.senseon.io

There is a hyphen before logs. The form with a dot (<code-name>.logs.senseon.io) does not resolve.

Method Port Use it for Availability
Syslog over TLS TCP 6514 Firewalls, switches, Linux hosts, anything that speaks syslog On request
JSON over HTTPS TCP 443 Log forwarders such as Fluent Bit, cloud functions, applications On by default
OpenTelemetry (OTLP) TCP 4317 (gRPC), 4318 (HTTP) Applications and collectors that export OpenTelemetry logs or metrics On request

These rules apply to all methods:

  • Connect by hostname. The endpoint routes connections on the TLS Server Name Indication (SNI). If a sender connects to an IP address or sends no SNI, the endpoint resets the connection. The sender gets no error that it can show you. If this rule prevents the use of a sender, refer to Devices that cannot send to a hostname.
  • TLS 1.2. If a sender can use only TLS 1.3, the handshake stops with an error.
  • Public certificate. The endpoint sends a Let's Encrypt certificate for *.senseon.io. A client certificate is not necessary.
  • No TLS inspection. If a proxy on the path inspects TLS or signs it again, the connection does not work. Exclude the hostname from TLS inspection.

SenseOn can also pull logs from Azure Event Hubs. With this method, no sender connects to the hostname. Refer to Azure and AWS.

💡 Note: If SenseOn onboarded your organisation on an earlier version of the platform, it is possible that you have an endpoint in the form <code-name>-collector.snson.net. Continue to use this endpoint until SenseOn asks you to move.

Requirements

Syslog over TLS

  • Send to <code-name>-logs.senseon.io, TCP 6514.
  • Use octet-counted framing, as RFC 5425 describes. With this method, each message has a length before it. The endpoint also accepts newline-delimited framing. But octet-counted framing is safer for messages that contain line breaks.
  • Start each message with a standard syslog header (RFC 5424 or RFC 3164). The header starts with the priority in angle brackets, for example <134>.

JSON over HTTPS

  • POST to https://<code-name>-logs.senseon.io/, the root path.
  • Send one JSON object, or a JSON array of objects, per request. Each array element becomes one log record.
  • Do not send newline-delimited JSON (NDJSON) with more than one object per request. The endpoint rejects it with HTTP 400.
  • If the endpoint accepts a request, it returns HTTP 200. The smallest valid log is any JSON object, for example {"event": "connection"}.

OpenTelemetry (OTLP)

  • Export to https://<code-name>-logs.senseon.io:4318 (HTTP) or <code-name>-logs.senseon.io:4317 (gRPC).
  • Send logs or metrics. The endpoint does not ingest traces.
  • For a worked example, refer to Claude Code Telemetry.

Syslog: Send the format of the device without changes, for example FortiGate key-value, CEF, LEEF or Cisco syslog. Do not convert it to JSON before you send it, because SenseOn recognises each product by its native format.

  • If the sender offers RFC 5424, use it. RFC 3164 also works.
  • Keep the header. A custom template that removes the <priority> header stops ingestion. The endpoint rejects the messages, or it closes the connection.
  • With rsyslog, use the built-in RSYSLOG_SyslogProtocol23Format (RFC 5424) or RSYSLOG_ForwardFormat templates. RSYSLOG_ForwardFormat is better for devices that write unusual timestamps.

JSON: Send one object for each event. Include an ISO 8601 UTC timestamp. Include a field that names the source, so that you can see the difference between sources in Hunt Lab. To send a batch, use an array.

[
  {"timestamp": "2025-10-14T15:30:45Z", "source": "nginx", "remote_addr": "203.0.113.42", "status": "200"},
  {"timestamp": "2025-10-14T15:30:46Z", "source": "nginx", "remote_addr": "203.0.113.42", "status": "404"}
]

SenseOn accepts and stores the logs from a new source before SenseOn maps their format. Send SenseOn a sanitised sample early. Then SenseOn can map the format before your full volume arrives.


Receive Mode

Your devices and forwarders send directly to the ingestion endpoint. SenseOn runs all the other parts. This is the simplest setup, and data flows within minutes. Receive mode is applicable to any source that can do one of these tasks:

  • Send syslog over TLS with SNI (for example FortiOS 7.4, rsyslog, syslog-ng, Fluent Bit and Vector).
  • POST JSON over HTTPS.
  • Export OTLP.

Testing Your Connection

Syslog over TLS: Run this test after SenseOn tells you that syslog is on for your organisation. Run the test from the same network path as your devices:

openssl s_client -connect <code-name>-logs.senseon.io:6514 \
  -servername <code-name>-logs.senseon.io \
  -verify_hostname <code-name>-logs.senseon.io -verify_return_error -brief </dev/null

If the endpoint works, the output shows Verification: OK and Verified peername: *.senseon.io. This test must use OpenSSL 1.1 or later. The LibreSSL build that comes with macOS does not have -verify_hostname.

JSON over HTTPS:

curl -X POST "https://<code-name>-logs.senseon.io/" \
  -H "Content-Type: application/json" \
  --data '{"timestamp": "2025-10-14T15:30:45Z", "source": "test", "message": "Connection test"}'

If the endpoint works, it returns 200.

For step-by-step guides, refer to Configuration Examples and Ingesting Syslog from Network Appliances.

Devices that cannot send to a hostname

Some senders can send only to an IP address, or they send no SNI. For example, FortiOS 7.2 was seen to do this. Some networks allow outbound rules by IP address only. On the shared ports, the endpoint resets the connections in both cases. There are two options:

  • A dedicated, IP-restricted port. This option is for a sender that supports TLS. Only the fixed public IP addresses that your logs leave from can get to the port. You tell SenseOn these IP addresses. If the sender cannot use the hostname, SenseOn sets up the port without the need for SNI. SenseOn then gives you the IP addresses and the port.

    Set the sender to do a check of the certificate against <code-name>-logs.senseon.io. Usually, you do this in a peer name or server name setting. The check must use the hostname, because the certificate does not contain an IP address. To request a dedicated port, contact SenseOn support.

  • A relay. A relay is a host in your network that forwards logs to SenseOn by hostname. A relay is also applicable to devices that cannot use TLS. Refer to Ingesting Syslog from Network Appliances.


Relay Mode

A collector in your environment collects logs and forwards them to the ingestion endpoint. Use relay mode for these sources and conditions:

  • Network devices that can send only plain syslog to an IP address.
  • Containerised workloads.
  • High volumes.
  • Logs that you want to filter or transform before they leave your network.
Environment Guide
Firewalls, switches, routers, VPN concentrators Ingesting Syslog from Network Appliances
Kubernetes (EKS, GKE, AKS) Kubernetes Configuration Guide
Linux hosts with rsyslog Syslog (rsyslog)
VMs with Fluent Bit FluentBit Log Forwarder
Cloud Run and other Google Cloud services Configuration Examples

When you forward logs with Fluent Bit, set Format json on the HTTP output. json_lines sends newline-delimited JSON, and the endpoint rejects this JSON with HTTP 400.


Retrieve Mode

SenseOn pulls logs from the location where you keep them.

Azure and AWS

Azure Event Hubs. Refer to Azure Event Hub.

AWS works in the opposite direction. A forwarder Lambda in your account pushes logs to your ingestion endpoint. Refer to AWS CloudTrail and CloudWatch.

Other sources

SenseOn can pull logs from API-only sources, legacy systems, object storage buckets (S3, GCS, Azure Blob) and other custom sources. For these sources, this is a premium service, and a professional services engagement is necessary. To discuss it, contact SenseOn.


Troubleshooting

Symptom Likely cause
Connection is reset straight away No SNI. The sender connects to an IP address, or its TLS library does not send the server name. On 6514 or OTLP, it is possible that the method is not on for your organisation yet.
Syslog connection drops, or nothing arrives, when messages flow Lines start without a <priority> header, usually because of a custom template
Handshake stops with a protocol version error The sender allows only TLS 1.3. Allow TLS 1.2.
openssl shows a certificate not issued by Let's Encrypt A proxy inspects TLS. Exclude the hostname from TLS inspection.
HTTP 400 on 443 The body is not valid JSON, or is newline-delimited JSON
cannot resolve or Domain name not found The hostname is incorrect, frequently with a dot in place of the hyphen
Timeouts Outbound 6514, 443, 4317 or 4318 is blocked. On a dedicated port, it is also possible that your logs leave from an address that SenseOn does not have

The FAQ gives more answers.

Need help?

Contact SenseOn support at support@senseon.io. Send this information:

  • Your organisation name
  • Your ingestion hostname
  • The method that you use
  • The log source and its version
  • Sanitised example logs
  • Any error messages
  • Your configuration files

Additional Documentation