That looks like a bug to me. Do you think you could write a short script to reproduce the steps you took in altering the RP, and file an issue on GitHub?
On Fri, Feb 10, 2017 at 1:27 AM, Davis Kirkendall < davis.e.kirkend...@gmail.com> wrote: > Hi Mark, > The pretty json version of the first 4 points is in the post below. > I restarted the db (there was a similar bug in previous influx versions > where a restart fixed it: https://github.com/influxdata/influxdb/issues/ > 6946), but that didn't help. > I also upgraded to 1.2.0 which also didn't change anything. > > However, I did notice that there are two shards that span the same time > range: > 298 default autogen 298 2014-01-06T00:00:00Z 2014-01-13T00:00:00Z > 2019-01-07T00:00:00Z > 390 default autogen 390 2013-10-28T00:00:00Z 2014-01-20T00:00:00Z > 2019-01-14T00:00:00Z > > The default retention policy was changed a few days ago (before the data > was written). Could there be a bug involving altering the retention policy > and having overlapping shards representing the same data? > > > > > Am Freitag, 10. Februar 2017 10:17:27 UTC+1 schrieb Davis Kirkendall: >> >> { >> "results": [ >> { >> "series": [ >> { >> "name": "metricId_1_intervalId_5", >> "columns": [ >> "time", >> "flags", >> "forecastOffset", >> "intervalId", >> "sensorId", >> "sign", >> "sourceId", >> "statId", >> "value" >> ], >> "values": [ >> [ >> 1389574800000000000, >> 0, >> null, >> "5", >> "132", >> "1", >> "315", >> "0", >> 152800 >> ], >> [ >> 1389574800000000000, >> null, >> null, >> "5", >> "132", >> "1", >> "315", >> "0", >> 152800 >> ], >> [ >> 1389575700000000000, >> 0, >> null, >> "5", >> "132", >> "1", >> "315", >> "0", >> 160000 >> ], >> [ >> 1389575700000000000, >> null, >> null, >> "5", >> "132", >> "1", >> "315", >> "0", >> 160000 >> ] >> ] >> } >> ] >> } >> ] >> } >> >> >> >> Am Donnerstag, 9. Februar 2017 22:23:02 UTC+1 schrieb Mark Rushakoff: >>> >>> The indentation is hard to follow in a variable width font. Can you post >>> the JSON output? (i.e. from the influx CLI run `format json` and then your >>> query `select * from "metricId_1_intervalId_5" where time >= '2014 >>> -01-13T01:00:00Z' and sourceId = '315' limit 10`. >>> >>> It looks like maybe the forecastOffset tag is 0 in half of the points >>> and absent in the other half, in which case I believe the system is >>> behaving as expected. If it isn't that, are you able to try running with >>> 1.2? >>> >>> On Thu, Feb 9, 2017 at 10:02 AM, Davis Kirkendall < >>> davis.e.k...@gmail.com> wrote: >>> >>>> I'm trying to track down some weird behavior in Influxdb 1.1 >>>> >>>> >>>> > select * from "metricId_1_intervalId_5" where time >= >>>> '2014-01-13T01:00:00Z' and sourceId = '315' limit 10 >>>> >>>> name: metricId_1_intervalId_5 >>>> ----------------------------- >>>> time flags forecastOffset intervalId sensorId >>>> sign sourceId statId value >>>> 1389574800000000000 0 5 132 >>>> 1 315 0 152800 >>>> 1389574800000000000 5 132 >>>> 1 315 0 152800 >>>> 1389575700000000000 0 5 132 >>>> 1 315 0 160000 >>>> 1389575700000000000 5 132 >>>> 1 315 0 160000 >>>> 1389576600000000000 0 5 132 >>>> 1 315 0 183200 >>>> 1389576600000000000 5 132 >>>> 1 315 0 183200 >>>> 1389577500000000000 0 5 132 >>>> 1 315 0 111200 >>>> 1389577500000000000 5 132 >>>> 1 315 0 111200 >>>> 1389578400000000000 0 5 132 >>>> 1 315 0 158400 >>>> 1389578400000000000 5 132 >>>> 1 315 0 158400 >>>> >>>> The interesting part is that "flags" and "values" are fields, meaning >>>> that there should be no way for duplicate points such as [1,2], [3,4] and >>>> so on to exist since timestamp and tag set are the same. >>>> >>>> > show field keys from metricId_1_intervalId_5 >>>> >>>> name: metricId_1_intervalId_5 >>>> ----------------------------- >>>> fieldKey fieldType >>>> value float >>>> flags float >>>> >>>> >>>> Am I missing something or has anyone experienced something comparable >>>> before? Could this be some sort of bug with the sharding process? >>>> >>>> >>>> -- >>>> Remember to include the version number! >>>> --- >>>> You received this message because you are subscribed to the Google >>>> Groups "InfluxData" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to influxdb+u...@googlegroups.com. >>>> To post to this group, send email to infl...@googlegroups.com. >>>> Visit this group at https://groups.google.com/group/influxdb. >>>> To view this discussion on the web visit https://groups.google.com/d/ms >>>> gid/influxdb/85f0e693-33d9-4fca-af79-8c1723d732bb%40googlegroups.com >>>> <https://groups.google.com/d/msgid/influxdb/85f0e693-33d9-4fca-af79-8c1723d732bb%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> For more options, visit https://groups.google.com/d/optout. >>>> >>> >>> -- > Remember to include the version number! > --- > You received this message because you are subscribed to the Google Groups > "InfluxData" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to influxdb+unsubscr...@googlegroups.com. > To post to this group, send email to influxdb@googlegroups.com. > Visit this group at https://groups.google.com/group/influxdb. > To view this discussion on the web visit https://groups.google.com/d/ > msgid/influxdb/2e723dc5-b175-462b-91f1-a80043098e12%40googlegroups.com > <https://groups.google.com/d/msgid/influxdb/2e723dc5-b175-462b-91f1-a80043098e12%40googlegroups.com?utm_medium=email&utm_source=footer> > . > > For more options, visit https://groups.google.com/d/optout. > -- Remember to include the version number! --- You received this message because you are subscribed to the Google Groups "InfluxData" group. To unsubscribe from this group and stop receiving emails from it, send an email to influxdb+unsubscr...@googlegroups.com. To post to this group, send email to influxdb@googlegroups.com. Visit this group at https://groups.google.com/group/influxdb. To view this discussion on the web visit https://groups.google.com/d/msgid/influxdb/CALxJwdM%3DopSQor4mEezErkeGotxeF8Q1LEe2f3FLhN7u2Xt72Q%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
