Sidecar fetch failed

[Joy91-Lin]

my testnet Asset Hub sidecar on Westend/Paseo stopped working and produced the logs below. After restarting the sidecar, it resumed normal operation.

Could you help us understand what might have caused this error, and whether there are any recommended solutions or workarounds for this issue?

• Environment / Versions

  • Paseo Asset Hub: 1.20.0-5ab401310d2 (Polkadot stable2509)
  • Paseo sidecar: v20.10.2
  • Westend Asset Hub: 1.19.2-2caeef482a4 (Polkadot polkadot-stable2506-2)
  • Westend sidecar: v20.7.3

• Log

Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 warn: API/INIT: Api will be available in a limited mode since the provider does not support subscriptions
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 error: RPC-CORE: getBlockHash(blockNumber?: BlockNumber): BlockHash:: fetch failed
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: Failed HTTP Request: {"method":"chain_getBlockHash","params":[0]}
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 error: API/INIT: Error: FATAL: Unable to initialize the API: fetch failed
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: Failed HTTP Request: {"method":"chain_getBlockHash","params":[0]}
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]:     at #onProviderConnect (/opt/substrate-api-sidecar/node_modules/@polkadot/api/cjs/base/Init.js:390:27)
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]:     at process.processTicksAndRejections (node:internal/process/task_queues:105:5)
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 error: RPC-CORE: getRuntimeVersion(at?: BlockHash): RuntimeVersion:: fetch failed
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: Failed HTTP Request: {"method":"state_getRuntimeVersion","params":[]}
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 error: RPC-CORE: chain(): Text:: fetch failed
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: Failed HTTP Request: {"method":"system_chain","params":[]}
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 error: RPC-CORE: properties(): ChainProperties:: fetch failed
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: Failed HTTP Request: {"method":"system_properties","params":[]}
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: 2026-02-04 06:57:56 error: RPC-CORE: methods(): RpcMethods:: fetch failed
Feb 04 06:57:56 kusuma-asset-hub01 yarn[988427]: Failed HTTP Request: {"method":"rpc_methods","params":[]}

[Imod7]

Based on the logs you provided, this appears to be a network or node connectivity issue rather than a problem with Sidecar itself. The fetch failed errors indicate that Sidecar was unable to reach the underlying node's RPC endpoint during initialization.
A few questions and recommendations:

• Node configuration: Are you running your own node, or connecting to a public RPC endpoint? If you are using a public endpoint, it may have experienced temporary downtime or rate limiting.
• Monitoring: If you are running your own node, do you have monitoring in place to detect any downtime issues? This would help identify the root cause of similar incidents in the future.

Since restarting Sidecar resolved the issue, I would recommend:

• Adding a retry logic for connection failures
• Adding a process to automatically restart Sidecar in case of initialization failures

[Joy91-Lin]

I am running my own full nodes and also have monitoring in place.

During this incident, the nodes themselves did not show any syncing issues. Only Sidecar went down. In addition, my KSM and Paseo nodes are running on separate VMs, but their Sidecar services failed at the same time.

Where should the retry connection logic be implemented?
Is there any existing configuration in Sidecar that allows us to adjust retry / timeout / backoff behavior for RPC connections?