Get Telegraf to send to multiple InfluxDB hosts

Telegraf
1

I totally get that I can use Enterprise Influx to solve this issue and understand if that is the only way to get what I want out of the box.

I have two servers that I want to use as influxdb host and I want my servers which are running telegraf daemons to post to both servers. I know that if I setup urls = [“server1”,“server2”] in [[outputs.influxdb]] I get round robin sending not duplicate sending.

Is there anyway short of placing a multiplexer in the path of telegraf to get a single telegraf daemon to send to multiple influx servers.

2

Yes, you can define multiple outputs:

[[outputs.influxdb]]
  urls = ["http://server1:8086"]
  # ... other options ...

[[outputs.influxdb]]
  urls = ["http://server2:8086"]
  # ... other options ...
3

This works great - Thanks very much for the response!

4

Hello Mr.Daniel
I try as you said.
But I can’t start Telegraf.

I try:

[[outputs.influxdb]]
  urls = ["http://server1:8086", "http://server2:8086"]

But Telegraf send metrics to server1 during period. Then Telegraf send metrics to server2 during period. And repeat again.

Please help me to fix it.
Thanks.

5

Just update your config to use the pattern shown in my comment above, with two independent influxdb outputs.

6

you need 2 separate output.influxdb definitions that each specify a different host. Not one output.influxdb that specifies 2 hosts.

7

@daniel
@aspitzer
I try as you said.
But I can’t start Telegraf.

8

Hi , why can’t you start Telegraf ? Does it give an error ?

9

Hi
I can’t start telegraf. This is log debug.

systemd[1]: telegraf.service: Main process exited, code=exited, status=1/FAILURE
systemd[1]: telegraf.service: Unit entered failed state.
systemd[1]: telegraf.service: Failed with result 'exit-code'.
systemd[1]: telegraf.service: Service hold-off time over, scheduling restart.
systemd[1]: Stopped The plugin-driven server agent for reporting metrics into InfluxDB.
systemd[1]: telegraf.service: Start request repeated too quickly.
systemd[1]: Failed to start The plugin-driven server agent for reporting metrics into InfluxDB.

Then, I comment and use only 1 output.influxdb. I started it success.

10

Thanks , can you give the
[[ outputs.influxdb]]
Sections of your configfile?
Is there any logging in the Telegraf logfile ?
Example : /var/log/telegraf/telegraf.log

Can you try to start Telegraf without using systemctl ?
Best regards

1 Like
11

Thank you very much ^^
I try do it and success ^^

12

Quick update – for InfluxDB Cloud 2, you’d want to have two sections that start with:

[[outputs.influxdb_v2]] # For InfluxDB 2.0

So, for example (be sure to scroll to get all the config lines):

# Configuration for sending metrics to InfluxDB 2.0 Open Source

[[outputs.influxdb_v2]] # For InfluxDB 2.0

## The URLs of the InfluxDB cluster nodes.
urls = ["http://127.0.0.1:9999"]

## Token for authentication.
token = "your_token"

## Organization is the name of the organization you wish to write to; must exist.
organization = "your_org"

## Destination bucket to write into.
bucket = "you_bucket"

# Configuration for sending metrics to InfluxDB 2.0 Cloud
[[outputs.influxdb_v2]] # For InfluxDB 2.0

## The URLs of the InfluxDB cluster nodes.
urls = ["https://us-west-2-1.aws.cloud2.influxdata.com"]

## Token for authentication.
token = "your_token"

## Organization is the name of the organization you wish to write to; must exist.
organization = "your_org" # This is your InfluxDB Cloud 2.0 org

## Destination bucket to write into.
bucket = "your_bucket"
13

Hello!

How I can use 2 hosts.
[agent]
interval = “10s”
debug = false
hostname = “host1”
round_interval = true
flush_interval = “10s”
flush_jitter = “0s”
collection_jitter = “0s”
metric_batch_size = 1000
metric_buffer_limit = 10000
quiet = false
logfile = “/var/log/telegraf/jitsi.log”
omit_hostname = false

And, I need one more host “host2” in the same telegraf.conf. Becouse I have 2 [[inputs.http]] from different hosts. And when I use “host1” I did not see “host2” metrics in Grafana