Collect debug info

What to collect before reporting a Curio issue.

When asking for help, posting the right info up front saves hours.

Copy/paste this checklist and fill it in.

Environment

• Curio version:

curio --version

• Deployment type:

  • systemd / docker / k8s / other

• Enabled layers (and where you set them):

  • /etc/curio.env (CURIO_LAYERS=...) or docker env / service unit

Config snapshot

If you use layered config, include the layer list and the relevant sections.

• If available in your build:

curio config view --layers <comma-separated-layers>

• Otherwise: paste the relevant layer TOML sections (Subsystems/Fees/Market/Ingest/HTTP/etc.).

Logs (include context)

• systemd:

journalctl -u curio -n 2000 --no-pager

• docker:

docker logs --tail 2000 <container>

Include:

• the exact error line

• ~30–100 lines before/after

DB connectivity and health

Confirm the Curio host can reach Yugabyte (YSQL):

ysqlsh -h "$CURIO_DB_HOST" -p "${CURIO_DB_PORT:-5433}" -U "$CURIO_DB_USER" -d "${CURIO_DB_NAME:-yugabyte}" -c "select 1;"

If you have multiple DB hosts, include the host list and whether load-balancing is enabled.

UI / task context

• Screenshot of failing task(s)

• The task ID(s) from the UI

Sealing issues

Include:

• sp_id (miner id)

• sector number(s)

• which stage (SDR/TreeD/PC1/PC2/C2/PreCommit/Commit/WdPoSt)

Market / deal ingestion issues

Include:

• deal UUID

• piece CID

• whether offline/online

• the URL you provided (if any) and headers (redact secrets)

• indexing status (IPNI enabled? CheckIndex errors?)

PDP issues

Include:

• PDP endpoint URL you are testing

• the command used (e.g. pdptool ping ...)

• whether you are testing locally or from outside the network

Redaction

Before posting publicly:

• redact API tokens, passwords, and private IPs (if needed)

• keep message CIDs / actor IDs (they’re useful for debugging)