Assets
Relationships, impact, and availability targets
Asset relationship tree, impact tracing, and monthly availability targets per monitor.
The asset detail page shows two complementary things: what depends on what (relationships and impact tracing), and how well a component meets its availability target (a monthly target per monitor). Both are visible only to users with the View assets permission. Monitoring and relationships do not appear in the requester portal.
Relationship tree
The Relationships card on the asset detail page has two views: List and Tree. Your last choice is remembered in this browser. Adding and removing relationships stays in the List view.
The Tree view has three sections:
| Section | Contents | Example wording |
|---|---|---|
| Upstream | What this asset needs, across levels | "depends on", "runs on", "part of" |
| Downstream | What relies on this asset, across levels | "used by", "runs", "contains" |
| Connected | Assets linked with "connected to", one level only | "connected to" |
Each row reads from parent to child. For example, in a VM's Downstream section, the row "used by APP-HR" means application APP-HR depends on that VM.
"Connected to" relationships are not traced further, because they do not say who depends on whom.
Assets with several parents and circular relationships
Each asset appears once in the tree, under its closest parent:
- Other parents are listed on that asset's row, for example "also depends on VM-02". Under that other parent a reference row reads "— already shown under VM-01". The Show button opens the original branch and moves focus there.
- Circular relationships (A depends on B and B depends on A) are marked "— circular, back to …". These rows cannot be expanded, so you never go round in circles.
- Asset counts in the summary are always unique. An asset with two parents is not counted twice.
The first level is visible right away; deeper branches open with the arrow button or Expand all. Branches with more than 20 rows are shortened with a Show all button. A collapsed branch that contains an asset observed DOWN gets a badge "n observed DOWN below". The tree stops at 10 levels or 500 assets, and the card tells you when that happens.
Trace impact
The Impact if disrupted card answers "if this asset goes down, what else is affected?". Click Trace impact to see:
- Potentially affected (from relations): the number of assets and services that rely on this asset, directly or across levels. This is an estimate from recorded relationships, not an actual disruption. This estimate does not account for redundancy groups, so a standby device may keep a service running. The figure that does is in Combined availability.
- Observed now (monitoring): this asset's own monitor status, and how many assets in the tree are observed DOWN or have no data.
- The list of services that may be disrupted, then the downstream asset tree.
The Observed DOWN and No data badges in the tree come only from that asset's own monitor. An asset without a badge is only potentially affected.
Monthly availability target
Targets are set per monitor, because availability is also calculated per monitor. An asset with several monitors (for example ping and HTTP) shows each monitor's target and result side by side, without averaging.
Setting a target
- Open the asset detail page and, on the Monitoring status card, click Edit on a monitor (or Add monitor).
- Enter Monthly availability target (%), for example
99.5. The value is 50–100 with at most three decimal places. The downtime budget for this month appears below the field. - Click Save.
A new target applies to the whole current month, so that month's result is recalculated using the monitor history already recorded. Earlier months keep being evaluated against their old target. Clearing the field removes the target from this month. Every change is recorded in Audit. Setting targets requires the Manage assets permission.
How it is calculated
The evaluation period is the calendar month in the workspace time zone, counted 24 × 7 without business hours. Planned maintenance that is reported as down still counts as down.
| Figure | Meaning |
|---|---|
| Actual | UP time divided by (UP + DOWN), with coverage next to it |
| Downtime budget | (100% − target) × the full length of the month. Example: 90% in a 30-day month = 3 days; 99.5% in a 31-day month = 3 h 43 min |
| Used | Observed DOWN time that month |
| Recorded budget left or exceeded | Budget minus used. When more has been used, the difference is shown as "exceeded". The budget left does not yet account for possible downtime during time without data, so the no-data time is always shown next to it |
| No data | Time without data: monitor not yet created, sender silent, monitor disabled. This time counts as neither up nor down |
The actual percentage uses known time only, while the target result uses the full-month budget and accounts for time without data. A percentage above the target therefore does not by itself mean Met.
Results
| Result | When |
|---|---|
| Provisional | The month has not ended. Figures run up to now and may still change |
| Missed | Downtime already exceeds the budget. This is final, even before the month ends |
| Met | The month has ended, and downtime plus time without data stays within the budget. The target is met even if all time without data was actually down |
| Cannot be determined | The month has ended, but time without data is larger than the remaining budget. There is not enough data to declare it met |
Insufficient data is never marked Met. In the month a monitor is created, the time before it existed counts as no data. That is why a partial first month usually cannot be determined, with the note "Monitored since …".
In the report
Open Monitoring → Report and choose This month or Last month. The table then gets a Target column, and a summary above it counts monitors per result. The column can be sorted (worst first) and filtered. The CSV download includes the columns target_persen, jatah_mati_detik, terpakai_detik, sisa_jatah_detik, terlampaui_detik, and penilaian_target. Other periods (7/30/90 days or a free date range) do not evaluate targets, because targets apply per calendar month.
A combined service level figure per service is not available yet.