Florian Boucault

I believe that the following query, a little simpler perhaps, exhibits the same issue: `groupByTags(divideSeriesLists(seriesByTag('name=random1'), seriesByTag('name=random2')), 'sum', 'name')` 

@fkaleo will review if robert is not able to atm

https://github.com/grafana/metrictank/pull/1897 might fix it, if not @robert-milan has another branch that should and will do a PR for

Could not figure out the cause yet.

I was able to reproduce it with query `diffSeries(summarize(groupByNode(stats.*.counters.statsd.metrics_received.count, 2, 'sum'), '1min', 'sum', false), summarize(timeShift(groupByNode(stats.*.counters.statsd.metrics_received.count, 2, 'sum'), '1d'), '1min', 'sum', false))` ``` PANIC: runtime error: index out of range [60]...

Even simpler reproduction query: `diffSeries(SINGLE_SERIES, summarize(timeShift(SINGLE_SERIES, '-1h'), '1min'))` Also breaks with `sumSeries`

Panic does not always happen, only with some time ranges.

`diffSeries(summarize(metrictank.stats.docker-env.default.tank.persist.values.rate32, '1min'), summarize(timeShift(metrictank.stats.docker-env.default.tank.persist.values.count32, '-1min'), '1min'))` recreates the issue. `summarize(metrictank.stats.docker-env.default.tank.persist.values.count32, '1min')` produces 2 data points `summarize(timeShift(metrictank.stats.docker-env.default.tank.persist.values.count32, '-1min'), '1min')` produces 3 data points

> `diffSeries(summarize(metrictank.stats.docker-env.default.tank.persist.values.rate32, '1min'), summarize(timeShift(metrictank.stats.docker-env.default.tank.persist.values.count32, '-1min'), '1min'))` recreates the issue. reversing the arguments to `diffSeries` and the panic seems gone

Series showing the issue with `from=1604708261266&to=1604708992552`