A meter monitoring recovery exercise observer record should document what happened, when it happened, who acted and which evidence supports the observation. Keep observed events separate from expected results and later explanations. Use a stated clock basis, mark uncertainty and record deviations without changing the original notes after the exercise. The observer does not approve the technical recovery merely by recording it. Their record supports the later review by preserving the sequence of decisions, dependencies and delays that a simple pass or fail result cannot explain.
A recovery exercise can involve the monitoring platform owner, site engineers, an integrator and a support provider. Each participant may see only part of the process. An observer captures the agreed scope and the sequence visible from their assigned position. They should not imply they verified systems or actions they could not inspect.
Decide whether the exercise is a discussion, a limited technical test or a broader operational rehearsal. State which recovery steps are simulated and which are executed. A tabletop discussion about restoring meter records does not establish that a backup can actually be restored. Conversely, a successful database restore does not prove that every operational role can execute a complete recovery plan.
The NIST test, training and exercise guide provides background on planning and evaluating exercises. The observer record suggested here is an engineering application of those ideas to external meter monitoring; it is not a mandatory form or a device feature.
Record the exercise identifier, objective, participating roles, approved environment and start conditions. Identify the external monitoring service and the relevant meter-data scope. Clarify whether production collection continues normally and which test environment is isolated. Do not allow an ambiguous exercise label to substitute for confirming the actual execution boundary.
State the success criteria before the exercise begins. These might include retrieving a selected archive, reading an agreed sample or confirming a documented dependency. The system owner sets the criteria using the recovery plan. The observer records whether evidence was presented for each criterion; they do not invent a threshold after seeing the result.
Use the existing historian restore acceptance guide for the technical checks. The observer record complements that acceptance evidence by showing who performed the steps, what blocked progress and which assumptions changed during the exercise.

Use a clearly stated time zone and identify the clock or event source used for observations. If several systems have different clock bases, record the known differences and the limits of correlation. Do not silently rewrite source timestamps to create an apparently precise sequence.
Distinguish an observer timestamp from a source event timestamp. The observer may note a screen result several seconds after the system produced it, or receive a verbal report much later. Record the source and timing basis when that difference matters. A statement such as “reported complete at this time” is more accurate than claiming direct observation of an earlier unseen event.
If timing is approximate, say so. The review can still learn that a missing approval caused a long wait without an invented exact duration. Avoid publishing detailed operational timestamps or identities outside the approved distribution group when the record contains sensitive information.
The following fields are qualitative recommendations. They do not specify a mandatory retention period or recovery-time target. Use the site’s approved objectives and documentation to populate those values where applicable.
| Event field | What to capture | What to keep separate |
|---|---|---|
| Observation time | Time and stated clock basis | Assumed time on another system |
| Actor or role | Who performed or reported the step | An unconfirmed person inferred from a login |
| Action | Specific observed or reported activity | A broad claim that recovery was complete |
| Evidence reference | Screen capture, log object or acceptance record | Uncontrolled copies of an entire archive |
| Expected outcome | Criterion agreed before testing | A new criterion invented during review |
| Actual outcome | What was visible and its limits | Root-cause speculation |
| Deviation | Missing input, unexpected state or changed sequence | Blame or an unsupported performance conclusion |
| Follow-up | Question, owner and review action | An undocumented promise to fix something |
A concise record is easier to evaluate than a transcript of every conversation. Capture decisions and events that affect the objectives, evidence quality, dependencies or safe execution. Keep the original notes available under the approved retention process even if a later report summarizes them.
The person directing the exercise may provide a simulated outage, missing credential or delayed supplier response. Mark these as exercise inputs, not real operational incidents. If the controller gives a participant information that would not normally be available, record that intervention because it affects how the outcome should be interpreted.
Similarly, identify skipped or simulated steps. A participant saying “we would contact the integrator” is evidence of an intended action, not evidence that a contact route worked. A discussion can reveal that the route is undocumented, but it cannot demonstrate the supplier’s actual availability or response time.
If the exercise must pause for safety or an unexpected production interaction, record the pause and the authorized restart decision. Do not treat a safe stop as an automatic failure of the recovery mechanism. The review should distinguish the technical objective from the need to control operational risk.
Record when a missing configuration file, license, account or network path prevents the next step. Identify who was expected to supply it and what workaround was used. Do not change the recovery method during observation merely to produce a passing result; any approved deviation should remain visible.
The historian restore dependency register is the appropriate place to maintain the technical dependency list. The observer record documents how that list performed during the exercise. A dependency discovered only during testing should become a tracked update after review, with an owner and verification step.
For the device boundary, consult the BY194 series product information and the exact ordered-model documentation. A recovery exercise for an external historian does not demonstrate internal backup, storage or restoration features in the instrument. Keep those product claims outside the observer’s conclusions unless directly supported.
Consider a hypothetical isolated recovery test in which the archive service starts successfully, but the team cannot interpret selected point identifiers because the approved mapping file is unavailable. The observer records the successful service start separately from the blocked sample interpretation. This example describes an external monitoring workflow, not a real customer event.
The exercise controller authorizes the team to obtain the mapping from a controlled repository. The observer records the intervention, source and resumed check. The technical reviewers can then determine whether the restored sample meets acceptance criteria while also noting that the original recovery package lacked a dependency.
The after-action report should not reduce this sequence to a simple unconditional pass. It can state that the sample was verified after an additional dependency was supplied, with an action to update the recovery package. A repeat test can then verify that the corrected package works without the same intervention.

After the exercise, distinguish observations, analysis and agreed actions. An observation describes what happened. Analysis proposes why it happened and identifies the supporting evidence. An action names the change, owner and verification method. Keeping these categories separate prevents a plausible explanation from becoming an undocumented fact.
Ask participants to correct factual errors without overwriting the original chronology. Record clarifications as dated additions linked to the relevant event. If two participants disagree, preserve the disagreement and identify the evidence needed to resolve it. Consensus is useful, but it cannot replace an unavailable log or acceptance record.
The NIST contingency planning guide provides broader background on maintaining recovery plans. Use the exercise review to improve the actual plan and evidence package. The commissioning witness sign-off article concerns a different acceptance context; an observer’s exercise notes should not be represented as that commissioning approval.
The following educational video illustrates how simultaneous human actions can produce unexpected sequences. It supports discussion of coordination and event order, while the exercise record must still rely on the actual observations and evidence.
No. It preserves observations and their context. Technical acceptance depends on the predefined criteria and supporting verification evidence, with any interventions or limitations stated explicitly.
No. Identify them as simulated or discussed. A participant describing a future action does not demonstrate that the corresponding system, contact route or procedure actually worked.
Yes, through traceable clarifications that preserve the original record. Record who supplied the correction, when it was added and which observation or evidence it affects.
A useful action identifies the problem, accountable owner, intended change and verification method. A general statement to improve the plan leaves the underlying dependency or coordination failure unresolved.