### Describe Your Changes
* check if `lastValue` was seen at least twice with different
timestamps. Otherwise, the difference between last timestamp and
previous timestamp could be `0` and will result into `NaN` calculation
* check if there items left in lastValue map after staleness cleanup.
Otherwise, `rate_avg` could have produce `NaN` result.
### Checklist
The following checks are **mandatory**:
- [x] My change adheres [VictoriaMetrics contributing
guidelines](https://docs.victoriametrics.com/contributing/).
---------
Signed-off-by: hagen1778 <roman@victoriametrics.com>
- Use bytesutil.InternString() instead of strings.Clone() for inputKey and outputKey in aggregatorpushSamples().
This should reduce string allocation rate, since strings can be re-used between aggrState flushes.
- Reduce memory allocations at dedupAggrShard by storing dedupAggrSample by value in the active series map.
- Remove duplicate call to bytesutil.InternBytes() at Deduplicator, since it is already called inside dedupAggr.pushSamples().
- Add missing string interning at rateAggrState.pushSamples().
Updates https://github.com/VictoriaMetrics/VictoriaMetrics/pull/6402
Added `rate` and `rate_avg` output
Resource usage is the same as for increase output, tested on a benchmark
---------
Signed-off-by: hagen1778 <roman@victoriametrics.com>
Co-authored-by: hagen1778 <roman@victoriametrics.com>
