Smart buildings
Temperature, CO2 and occupancy are stored per floor, with raw data for a day and hourly averages for reports.
Core dataIoTPerformance and Scale plans
IoT and Metrics Time Series
Ingest sensor and metrics samples, keep raw data for as long as you need, copy it to long-retention series, and aggregate into buckets on demand.
- per series
- Retention
- copy rules
- Auto
- to find series
- Labels
Step 1 of 6
Create a series
Keep raw samples for 24 hours and label them by building and type.
Example sessionplaying
# Create a series
hotkv> TS.CREATE temp:b4:f2 RETENTION 86400000 LABELS building b4 type temperature
OK
An animated walkthrough with real HotKV commands and example replies. Click any step to jump to it.
The problem
Thousands of sensors each send a reading every few seconds. Raw data is needed for a few hours for alerting, a year of history for reports, and finding the right series means searching by building, floor or device type. Running a separate time series database for that is one more system to operate.
How HotKV handles it
Time series in HotKV have a retention period and labels. Devices append samples with TS.ADD, a downsampling rule copies every new sample into a series you keep longer, and TS.AGG aggregates any range into fixed-width buckets with avg, min, max and other functions.
Step by step
What happens, command by command
- 01
Create a series
Keep raw samples for 24 hours and label them by building and type.
TS.CREATE temp:b4:f2 RETENTION 86400000 LABELS building b4 type temperature - 02
A long-term series
A second series that keeps samples for a year.
TS.CREATE temp:b4:f2:1y RETENTION 31536000000 - 03
Copy samples to it
Every new sample is also copied to the long-term series. Read it as hourly averages with TS.AGG.
TS.DOWNSAMPLE temp:b4:f2 temp:b4:f2:1y avg 3600000 - 04
Ingest samples
Sensors report through the gateway. * uses the server's time.
TS.ADD temp:b4:f2 * 21.4 - 05
Aggregate for a chart
Five-minute averages for the last hour, ready to plot.
TS.AGG temp:b4:f2 1790982000000 1790985600000 300000 avg - 06
Find series by label
Every temperature series in building b4.
TS.LIST LABEL building b4
Where it fits
Real-world scenarios
Fleet telemetry
Vehicle speed, battery and engine data are ingested continuously and charted per vehicle.
Application metrics
Request latency and error counts per endpoint feed live dashboards without a separate metrics database.
Built in
What you get
Retention per series
Old samples are removed automatically.
TS.CREATETS.RETENTIONDownsampling rules
New samples are copied to a long-retention series, which you aggregate when you read.
TS.DOWNSAMPLEBucket aggregation
Aggregate any range into fixed-width buckets.
TS.AGGTS.RANGELabels
Find series by label value.
TS.LISTTS.INFOCommon questions
Can I send my own timestamps?
Yes. Pass a Unix time in milliseconds instead of * to TS.ADD.
Which aggregations are available?
See the TS.AGG reference for the full list, including avg, min, max and sum.
Which plans include it?
Time series are part of the Performance and Scale plans.
Related use cases
Try it on your own workload
HotKV speaks the Redis protocol, so you can point an existing client at it and run these same commands.