evmRpcLatencyWatcher

Subscribe to RPC endpoints via WebSocket (eth_subscribe with newHeads) and measure when blocks arrive at the Portal versus when they appear at the RPC. Use this to monitor relative latency. The watcher is a query-transformer combo that already includes the block query it needs. Pass it directly as the source’s outputs:

import { evmPortalStream, evmRpcLatencyWatcher } from "@subsquid/pipes/evm";

const stream = evmPortalStream({
  id: "indexing-latency",
  portal: "https://portal.sqd.dev/datasets/base-mainnet",
  outputs: evmRpcLatencyWatcher({
    rpcUrl: ["https://base.drpc.org", "https://base-rpc.publicnode.com"],
  }),
});

for await (const { data } of stream) {
  for (const sample of data) {
    console.table(sample.rpc); // url, hash, receivedAt, portalDelayMs
  }
}

Parameters:

rpcUrl: Array of RPC WebSocket or HTTP URLs to compare against Portal
resolveTimeoutMs: How long to wait for an endpoint to report a block before recording it as rpc-behind. Defaults to 60_000

Output: Each batch carries a LatencySample[], holding the samples that became decidable in that batch, so iterate it. Each sample carries the observed block’s number and timestamp, portal.receivedAt (when the block arrived from the Portal), and an rpc array with one entry per configured endpoint. An rpc entry that resolved carries url, hash, receivedAt, and portalDelayMs. The delay is signed: a negative value means the Portal delivered the block before that endpoint did. An entry for an endpoint that did not report the block carries unresolved instead of receivedAt and portalDelayMs:

`unresolved`	Meaning
`rpc-behind`	The endpoint had not reached the block before the wait window closed, so the Portal is ahead by at least that window.
`rpc-missing`	The endpoint is already past the block but never reported it, after a reorg, a dropped update, or a block seen while the stream was backfilling.

Count unresolved entries rather than charting them. A missing delay is not zero.

Measured values include client-side network latency. Results are end-to-end delays as seen by the client, not pure Portal or RPC processing performance.

See the Data freshness monitoring guide for a complete example with Prometheus metrics.

Overview

Pipes SDK

Squid SDK

Alternative clients