Where communities thrive

  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
Repo info
Connor Peet
You could write a front door for influxdb that parses requests, validates permissions, and proxies to Influx
(Or PR some code directly into Influx! But that might be a little more challenging :) )
Sunil Shah
Hmm, thank you @connor4312. I'm afraid of some performance limitations, but I'll try that. I was hoping for an Influx ready solution because I see that being a recurring issue if someone is using this for IoT and has a lot of customer sited devices. That DB wide permission will allow users to write in each other's tags, albeit a small security concern it's still possible.
Connor Peet
Ah, yea. I saw someone else (or maybe it was you?) asking about that scenario a year ago or so
You could keep the performance impact pretty small I think if you write your Influx parser to be a streaming parser--the line protocol should be pretty friendly to that. Ideally that would keep the overhead a few microseconds + network latency. Maybe someone else has already written one you could take advantage of, or build off of :)
Sunil Shah
It wasn't me, I'm still pretty new to Influx. I'll investigate and post any progress. Thanks!
Gustavo Campos
I'm trying to create some CQ's to downsample old data, but I have a few tags in my series, and InfluxDB aggregate queries can't include tags, how do you approach that?
Jack Zampolin
@ssshah86 That feature will be available in the next release of InfluxEnterprise
@guhcampos You can only use functions on fields. Can you post an example of what you are trying to do over on community.influxdata.com?
Julien Syx

Hello everyone,
I'm stuck with a query :

> SELECT * FROM "flowerpower_in_greenhouse"
name: flowerpower_in_greenhouse
time            air_temp    battery    fertilizer    host        light    location    sensor    sensor_1        soil_moisture
1494700742000000000    0        100    0        serre-telegraf    0    serre            Flower_power_A4DA    22.1140003204346
1494700742000000000    0        92    0        serre-telegraf    0    serre            Flower_power_E945    14.5739994049072
1494700802000000000    0        100    0        serre-telegraf    0    serre            Flower_power_A4DA    22.1140003204346

Here are the series :


But > SELECT * FROM "flowerpower_in_greenhouse" WHERE "sensor" = 'Flower_power_E945' returns me nothing, any idea please ?

Also, here are the lines my exec input returns :
flowerpower,sensor=Flower_power_E945 light=0,soil_moisture=14.4289989471436,air_temp=0,fertilizer=0,battery=92
flowerpower,sensor=Flower_power_A4DA light=0,soil_moisture=22.1140003204346,air_temp=0,fertilizer=0,battery=100
Hi everyone
I'm trying to parse apache acces log, to get number of connection by IP
but I'm stuck with the query, I'm came from mysql, and how can I bypass the agregate/nonagregate query
I wan't to do that : SELECT COUNT(client_ip) AS nb, client_ip FROM apache_access_log GROUP BY client_ip
Jack Zampolin
@Seraf Sorry for not getting to you earlier! Can you post this question over on community.influxdata.com? Also it looks like you have the tags written as sensor_1 in the above example. Might be why you aren’t seeing results?
@T4lus You want count(distinct(client_ip)) AS visitors FROM apache_access_log
Also in the future if you have questions please direct them over to community.influxdata.com
Can anyone help me with installing influxdb on Ubuntu 17.04 (Zesty)?
@MaKin211 Only 2 steps.
  1. wget https://dl.influxdata.com/influxdb/releases/influxdb_1.2.4_amd64.deb
  2. sudo dpkg -i influxdb_1.2.4_amd64.deb
Hello! Is there a way in influxdb to get moving sum like moving avg function? I have the values [1,2,3,4,5,6,7] and want something like this select moving_sum(value, 2d) that returns [1, 3, 5, 7, 9, 11, 13] (or may be [3, 5, 7, 9 , 11, 13]). Something like this https://www.mathworks.com/help/matlab/ref/movsum.html
What is the best practice for downsampling to calendar month 1m? I need this for accounting.
Mukut Bhattacharjee
If i create a continuous query with groupby time interval of 5m, when will that CQ execute for the first time?
Mukut Bhattacharjee
Can anyone help me to understand how influx db chooses the default time buckets
Chris Evans
Hi team! I'm currently configuring InfluxDB and Kapacitor via a script which modifies the /etc/<program>/<program.conf> files with a sed -i command. Can anybody recommend a better, more reliable, future-forward option. I looked into using the HTTP API overrides but I couldn't find any examples of how to use it to override specific lines or config directives.
Faiaz Sanaulla
Hi guys, i'm runnig show queries command, but every time it return only the last one query. I'm using influxdb 1.3.2.
Paul Puschmann
@ChrisE-UoN I suggest using Ansible, Puppet, Chef or similar to configure InfluxDB.
(or manipulating files, config and packages)
Taimoor Chatha
I have a feeling my question either is something that is intuitive to db masters, or it's just not possibe, but I'll try it here as well: Is there any method to get unique values in influxdb and also get the timestamp? What's the reasoning behind the timestamp being epoch 0?
Taimoor Chatha
Is it possible to join the results that i get back with the distinct function, with the same table to get the timestamp?
Taimoor Chatha
select * from <measurement> group by <field> order by time desc limit 1 does what I want.
Calem Bendell
Heya everyone, is there a binary transfer format to insert over HTML?
Chris Dibble
If a node in my DOT graph is producing errors on kapacitor show task_name, is there any way to get more debug information on that error? Right now I just see a counter of the errors.
Hello! Is there a way to change field type of measurement? I inserted some values as integer, but now I need cast it to fload and insert float values.
Connor Peet
Not afaik
I'm running a phone server with events -- how would I go about logging these things in influxdb, if possible?
I understand how to log total active calls, per second.
But I want to log calls -- carrier, destination country, destination phone type, duration, etc.
Then, I want to see concurrent calls by country, by carrier, etc -- but these have a duration. Are durations possible with influxdb?
this room looks pretty quiet -- somewhere else more active?
support having?

"We are soon going to release Blockchain Tutorial using GOlang for Developers on our Gitter and telegram group - you can join our groups

Gitter : https://gitter.im/Karachain-Project
Telegram: https://t.me/karachain"

Ed Kim
Hi. Is there a way to store aggregated data only given the time range?
Let say we have millions of data points coming in every minute
Most of the time we only use mean value per minute
So we dont really need to store per second or millisecond data
So if there is a way to insert as it update the mean value of the minute range, we would save alot more computation time as well as the storage
Vitaly Zhuravlev
Hi, is there any good example how to use KapacitorLoopback node? https://docs.influxdata.com/kapacitor/v1.4/nodes/kapacitor_loopback_node/
Can it be used for chaining streams from different tick scripts?
Is there community/paid support? This room looks useless!!
Ed Kim
Haha agree
Hey just posting this here in case there's a chance somebody can help. I'm having trouble using GROUP BY inside subqueries. https://stackoverflow.com/questions/48653009/missing-data-using-group-by-with-influxdb-subqueries
nvm figured it out
tosheer kalra
Hey.. Have anyone tried pushing open tracing data in Jaeger or Zipkin format to influxdb and seen the traces using Jager / Zipkin as chronograf doesn't support it?