Redistimeseries

1) 1) (integer) 1548149180000
2) "26.199999999999999"
2) 1) (integer) 1548149185000
2) "27.399999999999999"
3) 1) (integer) 1548149190000
2) "24.800000000000001"
4) 1) (integer) 1548149195000
2) "23.199999999999999"
5) 1) (integer) 1548149200000
2) "25.199999999999999"
6) 1) (integer) 1548149205000
2) "28"
7) 1) (integer) 1548149210000
2) "20"

Integrations
RedisTimeSeries comes with several integrations into existing time series tools. One such integration is our [RedisTimeSeries adapter for Prometheus](https://github.com/RedisTimeSeries/prometheus-redistimeseries-adapter), which keeps all your monitoring metrics inside RedisTimeSeries while leveraging the entire Prometheus ecosystem.
Furthermore, we also created direct integrations for [Grafana](https://github.com/RedisTimeSeries/grafana-redistimeseries) and [Telegraph](https://github.com/RedisTimeSeries/telegraf). [This repository](https://github.com/RedisTimeSeries/prometheus-demos) contains a docker-compose setup of RedisTimeSeries, its remote write adaptor, Prometheus and Grafana. It also comes with a set of data generators and pre-built Grafana dashboards.

This is a maintenance release for RedisTimeSeries 1.6

Update urgency: MODERATE: Program an upgrade of the server, but it's not urgent.


Details
**Security and privacy:**
- 1061 Internode communications encryption: support passphrase for PEM files

**Bug fixes:**
- 1056 Return error when a shard is down (in 1.6.8 - an empty result was returned)

**Notes:**
- Minimal Redis version: 6.0.16

This is the General Availability release of RedisTimeSeries 1.6

Highlights

RedisTimeSeries 1.6 adds support for aggregating across multiple time series (multi-key). Before this version, queries such as “the maximum observed value of a set of time series” needed to be calculated client-side. Such queries can now be computed server-side, leveraging the heart of RedisGears ([LibMR](https://github.com/RedisGears/LibMR)) for clustered databases.

This is the second Release Candidate for RedisTimeSeries 1.6

Details

**Features:**
- 882 RoF support
- 841 TS.DEL now updates all down-sampled series

**Performance enhancements:**
- 852 Improved queries performance: using the [Dragonbox](https://github.com/jk-jeon/dragonbox) algorithm
- 866 Memory reduction: [LibMR](https://github.com/RedisGears/LibMR) integration
- 925 Improved replication speed: it is now possible to enable short-read on all redis-server versions
- 922 Improved decompression performance

- Introduction of `GROUPBY` and `REDUCE` in [TS.MRANGE and TS.MREVRANGE](https://oss.redis.com/redistimeseries/commands/#tsmrangetsmrevrange) to add support for "multi-key aggregation" and support for such aggregations spanning multiple shards, leveraging [LibMR](https://github.com/RedisGears/LibMR)). Currently, we support `min`, `max`, and `sum` as reducers and grouping by a label.

- 722, 275 Filter results using `FILTER_BY_TS` by providing a list of timestamps and `FILTER_BY_VALUE` by providing a `min` and a `max` value ([TS.RANGE, TS.REVRANGE](https://oss.redis.com/redistimeseries/commands/#tsrangetsrevrange), [TS.MRANGE, and TS.MREVRANGE](https://oss.redis.com/redistimeseries/commands/#tsmrangetsmrevrange))

- 603, 611, 841 Introduction of [TS.DEL](https://oss.redis.com/redistimeseries/commands/#tsdel) which allows deleting samples in a time series within two timestamps (inclusive).

- 762 Limit the number of returned labels in the response of read commands ([TS.MRANGE, TS.MREVRANGE](https://oss.redis.com/redistimeseries/commands/#tsmrangetsmrevrange), and [TS.MGET](https://oss.redis.com/redistimeseries/commands/#tsmget)) using `SELECTED_LABELS`. This can be a significant performance improvement when returning a large number of series.

- 655, 801 Ability to align the aggregation buckets with the requested start, end, or specific timestamp on aggregation queries using `ALIGN` ([TS.RANGE, TS.REVRANGE](https://oss.redis.com/redistimeseries/commands/#tsrangetsrevrange), [TS.MRANGE, and TS.MREVRANGE](https://oss.redis.com/redistimeseries/commands/#tsmrangetsmrevrange)).

- 675 Add keyspace notifications for all CRUD commands. Check out [this test](https://github.com/RedisTimeSeries/RedisTimeSeries/blob/master/tests/flow/test_ts_keyspace.py) for the details.

Details

This is the first Release Candidate for RedisTimeSeries 1.6

Headlines

This new major version adds support for reducing different time series (multi-key). Prior to this version a query for “the maximum observed value of a set of time series” needed to be calculated client-side. This is now accelerated and computed server-side, leveraging [RedisGears](redisgears.io) for clustered databases.

Details

**Features:**
- Introduction of [`GROUPBY` and `REDUCE` in `TS.MRANGE`](https://oss.redis.com/redistimeseries/master/commands/#tsmrangetsmrevrange) to add support for "multi-key aggregation" and support for such aggregations spanning multiple shards, leveraging RedisGears. Currently, we support min, max and sum as reducers and grouping by a label.
- 722,275 Filter results using [`FILTER_BY_TS`](https://oss.redis.com/redistimeseries/master/commands/#tsrangetsrevrange) by providing a list of timestamps and [`FILTER_BY_VALUE`](https://oss.redis.com/redistimeseries/master/commands/#tsrangetsrevrange) by providing a min and a max value.
- 603, 611 Introduction of [`TS.DEL`](https://oss.redis.com/redistimeseries/master/commands/#tsdel) which allows deleting samples in a time series within two timestamps (inclusive).
- 762 Limit the number of returned labels in the response of read commands using `SELECTED_LABELS`. This can be a significant performance improvement when returning a large number of series.
- 655, 801 Ability to [align the aggregation buckets](https://github.com/RedisTimeSeries/RedisTimeSeries/issues/655#issuecomment-831865772) with the requested start, end or specific timestamp on [aggregation queries](https://oss.redis.com/redistimeseries/master/commands/#tsrangetsrevrange).
- 675 Add keyspace notifications for all CRUD commands. Check out [this test](https://github.com/RedisTimeSeries/RedisTimeSeries/blob/master/tests/flow/test_ts_keyspace.py) for the details.

Notes:
- The version inside Redis will be 1.6.0 in semantic versioning. Since the version of a module in Redis is numeric, we could not add a Release Candidate flag. The GA version will follow soon.
- There is no need to load RedisGears in a single shard database. In clustered databases, however, the use of `GROUPBY` will require [RedisGears](redisgears.io) to be loaded in Redis as well. In the future GA version, we plan to embed parts of RedisGears as a library in RedisTimeSeries to simplify this. In this case, the embedded RedisGears won't expose any programmable options.