Socket

Listen on address and parse framed payloads as records. Use it for simple network producers that are not HTTP, Kafka, or a cloud queue.

Skippr, your data department

Speak to an expert. Upfront insights in days, not months.

Book an intro call →

20 minutes · No sales pitch

Before you begin

• A free bind address on the Skipprd host (0.0.0.0:9000 or similar).
• A producer that can connect and send the framing you choose.

Configure

Python

from skippr import Config, DataSourceSocket, Pipeline

cfg = Config.discover()
src = cfg.data_source(
    "net",
    DataSourceSocket(mode="tcp", address="0.0.0.0:9000", framing="line", format="json"),
)
cfg.pipeline("net", Pipeline(data_source=src))
cfg.save()

CLI

skipprd connect data-source socket \
  --pipeline net \
  --name net \
  --mode tcp \
  --address 0.0.0.0:9000 \
  --framing line \
  --format json

YAML

data_sources:
  net:
    Socket:
      mode: tcp
      address: "0.0.0.0:9000"
      framing: line
      format: json

Run skipprd sync (not --once) so the listener stays up.

Options

Key	Type	Required/Default	Description
mode	string	Required	tcp or udp
address	string	Required	Bind address host:port
framing	string	Not set	How messages are split (line, …)
format	string	Not set	Payload format
batch_size_bytes	integer	Not set	Max bytes per batch
batch_size_seconds	integer	Not set	Max seconds per batch

What gets synced

Each framed message becomes a record. Push sources can lose a message that arrived after accept and before the WAL commit — have producers retry. See Exactly-once delivery.

Troubleshooting

Symptom	Fix
Address already in use	Pick another port, or stop the other listener
No rows	Confirm the producer uses the same mode, port, and framing
Parse errors	Match format to the payload

Next steps

• HTTP server
• Kafka