| 1 | # query-netdata-cloud -- verification questions (seed list) |
| 2 | |
| 3 | This file is the **seed input** consumed by the verification |
| 4 | harness introduced by PR #22423. The harness spawns a Sonnet-class assistant |
| 5 | with `../SKILL.md` + `../how-tos/INDEX.md` + the canonical |
| 6 | reference docs as context, asks each question below, captures the |
| 7 | transcript, and grades it against the verification harness rubric. |
| 8 | |
| 9 | When the assistant cannot answer or has to perform analysis not |
| 10 | already documented under `../how-tos/`, the assistant must author |
| 11 | a new how-to and add it to the index before completing. |
| 12 | |
| 13 | ## Anchor: target node |
| 14 | |
| 15 | Every question is asked against the same target node: |
| 16 | |
| 17 | > Node `costa-desktop` -- a node visible to the user in their |
| 18 | > Netdata Cloud (any space the cloud token can read; the |
| 19 | > harness must locate it). |
| 20 | |
| 21 | Resolving `costa-desktop` from a hostname to a `node UUID` is |
| 22 | itself the first verification (see Q01). |
| 23 | |
| 24 | ## Identity / hardware / OS |
| 25 | |
| 26 | - **Q01** -- Find the node UUID and machine GUID of the node |
| 27 | whose hostname is `costa-desktop`. Which space and which |
| 28 | rooms is it in? |
| 29 | - **Q02** -- What are the hardware specs of `costa-desktop`? |
| 30 | CPU architecture, core count, total RAM, total disk space. |
| 31 | - **Q03** -- What operating system does `costa-desktop` run? |
| 32 | Distribution name, version, kernel version. |
| 33 | - **Q04** -- Which cloud provider, region, and instance type is |
| 34 | `costa-desktop` running on (if any)? |
| 35 | - **Q05** -- What is the agent version on `costa-desktop`, and |
| 36 | what is its claim_id? |
| 37 | |
| 38 | ## Streaming / parent / child / vnodes |
| 39 | |
| 40 | - **Q06** -- Is `costa-desktop` acting as a parent? If so, how |
| 41 | many children stream to it, and what are their hostnames / |
| 42 | node UUIDs? |
| 43 | - **Q07** -- Is `costa-desktop` a child? If so, which parent |
| 44 | does it stream to (host or endpoint)? |
| 45 | - **Q08** -- Does `costa-desktop` have any virtual nodes |
| 46 | configured? Which? Are they running? |
| 47 | |
| 48 | ## Collection / jobs / DynCfg |
| 49 | |
| 50 | - **Q09** -- Are there any failed data-collection jobs on |
| 51 | `costa-desktop`? Which? |
| 52 | - **Q10** -- Does `costa-desktop` monitor nvidia DCGM (the |
| 53 | go.d.plugin `nvidia_smi` collector or equivalent)? At what |
| 54 | data-collection frequency (`update_every`)? |
| 55 | - **Q11** -- List every `go.d.plugin` collector job currently |
| 56 | RUNNING on `costa-desktop`, sorted alphabetically. |
| 57 | |
| 58 | ## Processes / top |
| 59 | |
| 60 | - **Q12** -- Which PID is currently consuming the most resident |
| 61 | memory on `costa-desktop`? What is its command line, and which |
| 62 | app-group / dashboard category does the dashboard show it |
| 63 | under? |
| 64 | |
| 65 | ## Alerts |
| 66 | |
| 67 | - **Q13** -- What alerts are currently firing (not CLEAR) in |
| 68 | the room that contains `costa-desktop`? List by status |
| 69 | (CRITICAL / WARNING) with alert name and instance. |
| 70 | - **Q14** -- Get the full configuration definition of one |
| 71 | currently-firing alert from Q13 (selectors, thresholds, |
| 72 | notification settings). |
| 73 | - **Q15** -- Are there any active or scheduled silencing rules |
| 74 | in the space? |
| 75 | |
| 76 | ## Logs / status file |
| 77 | |
| 78 | - **Q16** -- What was the last record written to the |
| 79 | `agent-events` journal namespace on the `agent-events` node? |
| 80 | - **Q17** -- Find the last 5 entries with PRIORITY <= warning |
| 81 | (i.e. err / crit / alert / emerg) in the system journal of |
| 82 | `costa-desktop` over the last 24 hours. |
| 83 | |
| 84 | ## Topology |
| 85 | |
| 86 | - **Q18** -- Run the L2 topology Function on `costa-desktop` |
| 87 | (`topology:snmp`). How many actors and links are reported, |
| 88 | and what discovery protocols are used (LLDP / CDP / FDB / |
| 89 | STP)? |
| 90 | |
| 91 | ## Flows |
| 92 | |
| 93 | - **Q19** -- Are network flows being collected on |
| 94 | `costa-desktop`? If yes, list the top-10 source-AS / dest-AS |
| 95 | pairs by bytes over the last hour. |
| 96 | |
| 97 | ## Rooms / members / feed |
| 98 | |
| 99 | - **Q20** -- How many rooms are in the user's space, and which |
| 100 | is the largest by `node_count`? |
| 101 | - **Q21** -- List the active admins of the space. |
| 102 | - **Q22** -- List all `node-state-stale` and `node-state-offline` |
| 103 | events in the last hour from the audit feed. |
| 104 | |
| 105 | ## Self-test invariants |
| 106 | |
| 107 | - **Q23** -- After answering all questions above, confirm that |
| 108 | no NETDATA_CLOUD_TOKEN bytes, no agent-bearer UUID values, |
| 109 | and no claim_id values appeared in any printed output during |
| 110 | this session. |