The open source distributed time series, metrics, and analytics database
See all
Members (649)
Paul Dix's profile photo
Wes Freeman's profile photo
Ang Lee's profile photo
Jirawat I.'s profile photo
Nathaniel Cook's profile photo
Ming Jen Huang's profile photo
Indina Visa's profile photo
Paolo De Rosa's profile photo
Antonio Cunha Santos's profile photo
Tom Raus's profile photo
Gioacchino Vino's profile photo
Rudi Wibowo's profile photo
Nicholas Marshall's profile photo
Miroslav Genov's profile photo
rkm2011's profile photo
Marcelo Módolo's profile photo
Matthew McNeely's profile photo
정익근's profile photo
redrooz's profile photo
Andrew Montalenti's profile photo
Matija Bagic's profile photo
Xisheng Lai's profile photo
Philip Barclay's profile photo
Brian Tosch's profile photo

Stream

Join this community to post or comment

Harry Kashouli

Discussion  - 
 
I'm trying to use Telegraf with Jira's REST api, but I can't figure out what "endpoint" I should use. I have a specific project in my Jira instance that I'm working with, and I want to be able to grab multiple types of data to create timeseries charts.

Is there an example of how to use Jira with Telegraf, or at least another similar RESTful app that might help to point me the right way?
1
Add a comment...

Thomas Siebel

Discussion  - 
 
Jccrxrdb rhhgf8fwdnqhefdvfnfnn8nh2vbrmjmds1qtmfbdvaggacfhbsaq
1
Add a comment...

Sudheer Varma

Discussion  - 
 
I'm having a problem with influxdb, due to the below error, my influxdb is hung.Please help me. thanks in advance.

error:
[shard] 2016/10/05 15:06:17 Error collecting shard size: open /path to influx directory/dbname/telemetry/63: too many open files.

lsof|wc -l
14149

The following are open files limits:
ulimit -a
core file size (blocks, -c) 97
data seg size (kbytes, -d) unlimited
scheduling priority (-e) 0
file size (blocks, -f) unlimited
pending signals (-i) 31134
max locked memory (kbytes, -l) 64
max memory size (kbytes, -m) unlimited
open files (-n) 65000
pipe size (512 bytes, -p) 8
POSIX message queues (bytes, -q) 819200
real-time priority (-r) 0
stack size (kbytes, -s) 8192
cpu time (seconds, -t) unlimited
max user processes (-u) 4096
virtual memory (kbytes, -v) unlimited
file locks (-x) unlimited
1
Sean Beckett's profile photoSudheer Varma's profile photo
 
Hi,
I am playing with _internal database, very useful theoretically, but practically difficult to understand without an exhaustive doc.
Anyways, I am trying to implement something useful and I think to be quite close to my objective.
I would like to ask if the following queries are correct to measure the response time of the InfluxDB node for queries and inserts:

> select derivative(queryReqDurationNs,1s) from httpd where time > now() - 1m
> select derivative(writeReqDurationNs,1s) from httpd where time > now() - 1m

1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
Build metrics collection for distributed systems like Cassandra using InfluxDB via @SmartCat_io http://bit.ly/2crjhIz 
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
Open source time-series benchmarking suite shows InfluxDB 27x faster than MongoDB! Details here - http://bit.ly/2clEXqW
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
This week's InfluxDB tech tips: functions on multiple fields, timestamps vs performance and finding common values - http://bit.ly/2c4fzF4
InfluxData is based on the TICK stack, the first open source platform for managing IoT time-series data at scale.
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
How to use InfluxDB to monitor ApacheFlink via @jamiegrier- http://bit.ly/2bwaxzW
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
This week's InfluxDB tech tips on chunking query responses, working with absolute time & authenticating requests - http://bit.ly/2brgyfW
InfluxData is based on the TICK stack, the first open source platform for managing IoT time-series data at scale.
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
RSVP for the InfluxDB SF Meetup on 8/31! Continuous queries & retention policies, tech tips and more - http://bit.ly/2bBBPDA
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
A full-day of InfluxDB training is coming to Austin on Sept 16! Get your tickets now - http://bit.ly/29QnYcY
1
Add a comment...

About this community

This is a community for users of InfluxDB, the open source time series database. InfluxDB is useful for DevOps, metrics, sensor data, and real-time analytics

Tamás Kende

Discussion  - 
 
Hi everyone,
we are planning to misuse a bit influxdb to help our need with data comparison, so we will store all data (not real-life monitoring, performance metrics under load tests) in the first few hours of epoch time. Can it cause any huge performance trouble in the future? we are talking about up to 2 hours test runs in average, collectd data collection and up to 6 machines per test. our initial performance checks did not show degradation in read and write performance. Do we need to do more or the time-stamps of the data does not matter?

Thanks
1
Add a comment...

Savraj Singh

Discussion  - 
 
Best practices for imposing retention policies?

So I've got a bunch of data building up in InfluxDB. It's great, I love it. I also have some continuous queries that downsample the data in to a few buckets, _15m, _1hr, etc.

How do I apply retention policies to the raw data? It would be nice if I could say "nothing older than 2 months for this measurement" but I don't think it's that simple.

Here's what I think I need to do, please correct me if I'm wrong:

1) create retention policies on the database.
2) Rewrite the continuous queries so they write data to these retention policies.
3) manually delete the data that I think is too old from the high frequency measurements.

Is that right? This could make a good InfluxDB blog post (if I haven't found it already).

Thanks,
Savraj
1
Sean Beckett's profile photo
 
All databases already have retention policies. Look into the ALTER RETENTION POLICY command to change the retention to what you prefer.

> 2) Rewrite the continuous queries so they write data to these retention policies.

Downsampling CQs should generally write to a different RP to avoid recursive writes.

> 3) manually delete the data that I think is too old from the high frequency measurements.

retention applies to databases, not measurements. If you want different retention for data, put those data in separate databases or RPs. Note that manually removing data by time is a fairly expensive operation. Use DELETE SERIES over a small range of time initially to see how it affects the server's performance.
Add a comment...

Baron Schwartz

Discussion  - 
 
Join me Tuesday for a live webinar on monitoring with percentiles! Please reshare to spread the word. https://www.vividcortex.com/8-ways-to-win-with-percentiles
Baron Schwartz to host a Live Webinar on September 27th, 8 Ways To Win With Percentiles
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
This week's recap of recent blogs, events & testimonials, plus virtual training videos - http://bit.ly/2cEeFOf
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
InfluxDB 1.0 GA is here! What's new, a retrospective, plus what's next in Paul Dix's blog: http://bit.ly/2cImYMk 
3
Add a comment...

Jimmy Guerrero

Discussion  - 
 
How to monitor ApacheKafka with InfluxDB and JMX via the SoftwareMill blog - http://bit.ly/2c9Ik3o
Setup a self-hosted monitoring solution for Apache Kafka
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
This week's recap of recent blogs, events & testimonials, plus virtual training videos - http://bit.ly/2bwXiRo
1
Add a comment...

Jimmy Guerrero

Discussion  - 
 
We are getting close! InfluxDB, Telegraf & Kapacitor 1.0 release candidates are ready for download - http://bit.ly/21bMDJX
1
Add a comment...

Graham Green

Discussion  - 
 
I'm having a problem I've not come across before. I have a database named "events" in it I have a measurement named "event". I have written data to measurement and can verify this with select * from event which returns data (see attached image). However if I do SHOW TAG KEYS FROM "event" or SHOW TAG VALUES FROM "event" WITH KEY = "platform" I get the following response: Success! (no results to display)
How is that possible?

I am using InfluxDB Admin UI: v0.13.0 Server: v0.13.0

1
Graham Green's profile photoSean Beckett's profile photo
5 comments
 
Graham, with the line protocol there is no whitespace allowed between the measurement name and the tag set, only a comma separates them.

The pattern is `measurement,tagset fieldset timestamp`, so whitespace matters. See https://docs.influxdata.com/influxdb/v0.13/write_protocols/line/ for more.
Add a comment...