Skip to main content

Time Series Analyzer (7801)

This is a tool for plotting the metrics held by the Time Series Engine (TSE). Metrics for the platform, edge, JVM, and the engine itself are gathered in a single tree, where you can select them, overlay them, and switch to a table or TQL view.

ItemValue
Address (TLS)https://<server-ip>:7801/
LoginTime Series Engine account
Other access pathManagement Console → Tools screen, Timeseries Engine card
This is a "look under the hood" tool, not a dashboard

For building field screens, use Canvas or PlantPulse Dashboard. This tool is for digging into what is happening right now — when a value spiked, or whether two metrics move together.


Screen​

Time Series Analyzer — screen with two JVM memory metrics selected and overlaid from the metric tree

No.What
①Time Range — presets like "Last 24 hours" and direct specification. The "Auto" option next to it enables periodic refresh
②Metric Search — search by name among 667 metrics
③Metric Tree — PLANTPULSE followed by EDGE · JMX · TSE. Clicking a leaf adds it to the chart (does not replace the current view)
④Chart Tools — save image · download · align axes · summarize · threshold lines · point/line · log axis · fullscreen
⑤Chart — view overlaid series together
⑥Time Brush — drag the lower band to narrow the interval. The chart above follows
⑦Series List — for each series: minimum · maximum · average · most recent. Use the eye icon to temporarily hide, × to remove
⑧View Switcher — 테이블 · JSON · TQL. The bottom left shows the metric count, query time, and sample count

How to Read the Tree​

BranchContents
EDGEMetrics sent by edge gateways
JMXJVM process metrics
TSEEngine itself — DATASTORE · HTTP · INGEST_EXECUTOR · JVM · PROTOCOL · QUEUE · METRIC_COUNTERS

The fully qualified metric name is the branch path concatenated — e.g., PLANTPULSE.TSE.JVM.USED_MEMORY. Since that name appears verbatim in the series list, you can copy it from here to use in TQL or PlantPulse Dashboard.

The point is overlaying

Looking at USED_MEMORY alone, you can't tell if the sawtooth is normal. Adding MAX_MEMORY shows at a glance "how close to the ceiling" it is. The screen above is an example of that.


Table · JSON · TQL​

Table — view the plotted data as values. Use this when you need to confirm exactly when a point occurred.

Table view

TQL — shows the query behind the current screen. Copy it as-is for use in Data Gateway Console TQL mode or in a program.

TQL view

You can take what you built on screen directly

A query you construct by clicking is immediately a TQL statement. It's faster to build here and copy it out than to learn the syntax by hand.


When It's Slow​

First check the query time and sample count at the bottom left. The screen above read 158,484 samples in 1,132 ms. If it's slow, it's usually one of these two.

SymptomWhere to look
Sample count exceeds hundreds of thousandsNarrow the time range or cut the interval with the brush
Many series overlaidRemove the ones you're not looking at with × — hiding (eye icon) does not reduce the query

See Also​

Data Gateway Console (5501) · timeseries component · System Monitoring