Right here’s the Regular Availability originate for RedisTimeSeries 1.4.
Capability to backfill time series! You might presumably presumably additionally now add samples to a time series where the time of the sample is older than the most modern sample in the series. This enables:
- Adding out of present of samples to time series.
- Batch loading of historical samples into an existing series.
- Updating existing samples (for instance for compliance causes).
This has been the most requested characteristic for RedisTimeSeries. We glimpse ahead to your feedback so we can high-tail to a standard availability originate soon.
- Added functionality:
TS.MREVRANGE[commands] enable for querying in descending present of Timestamps.(https://oss.redislabs.com/redistimeseries/1.4/instructions/#tsmrangetsmrevrange)
- #503 – RDB saves the entire chunk as an different of particular particular person samples giving a bustle and explain enchancment when saving or loading an RDB file
- #502 – The skill to space, at introduction time, the knowledge half size of every chunk using flag
CHUNK_SIZE. TS.INFO uses
chunkSizeas an different of
- #437 Enable backfilling of samples and updating of existing samples
- Works with compressed and uncompressed series.
- This comes with a performance hit when a sample is written out-of-present. We can put up numbers when we’re most incessantly readily accessible, nonetheless are quiet concerned with optimisations.
DUPLICATE_POLICYpermits to configure on module, series and sample stage clear-reduce how one can address duplicate samples. A duplicate sample is a sample for which the series holds already a sample on the same timestamp. Indicate that the default behaviour is equal to v1.2:
The version internal Redis might be 10405 or 1.4.5 in semantic versioning. For the reason that version of a module in Redis is numeric, we might presumably presumably now not add an GA flag.