Annotations for the InfluxDB module

[fntlnz]

Signed-off-by: Lorenzo Fontana lo@linux.com

What this PR does / why we need it:
This PR adds support for NGINX InfluxDB Module to the Nginx Ingress controller. The PR to update the image is #2533

This PR allows to configure the InfluxDB module using annotations.

This is useful for those who want to monitor every request passing through the ingress controller by sending them to InfluxDB or to one of the supported output of Telegraf (e.g Kafka, graylog, mqtt, opentsdb and all the others)

Special notes for your reviewer:

• Please provide any hints! Need to know if I missed something
• Still testing if all I made here works, please poiint out anything!

[aledbf]

@fntlnz thank you for your contribution. Please split this PR in two. The first one updating build.sh and Makefile from the nginx directory and another adding the feature and updating the Makefile to use the new nginx image. Also please add e2e tests to verify the new feature

[fntlnz]

Ok @aledbf I moved the image modifications to #2533

I also generated bindata, I don't know if I had to, just let me know. Not needed, after a rebase the bindata was not needed anymore.

[codecov-io]

Codecov Report

Merging #2505 into master will decrease coverage by 0.07%.
The diff coverage is 33.89%.

@@            Coverage Diff             @@
##           master    #2505      +/-   ##
==========================================
- Coverage    40.7%   40.62%   -0.08%     
==========================================
  Files          73       74       +1     
  Lines        5034     5093      +59     
==========================================
+ Hits         2049     2069      +20     
- Misses       2711     2743      +32     
- Partials      274      281       +7

Impacted Files	Coverage Δ
internal/ingress/controller/config/config.go	98.23% <ø> (ø)	⬆️
internal/ingress/types_equals.go	12.41% <0%> (-0.09%)	⬇️
internal/ingress/controller/controller.go	2.23% <0%> (-0.02%)	⬇️
internal/ingress/annotations/annotations.go	80.95% <100%> (+0.3%)	⬆️
internal/ingress/controller/template/template.go	63.93% <28.57%> (-1.05%)	⬇️
internal/ingress/annotations/influxdb/main.go	38.46% <38.46%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update aa256ac...c3b896d. Read the comment docs.

[ElvinEfendi]

@fntlnz ingress-nginx already has /metrics endpoint where it exposes Prometheus metrics. I wonder why did you not use https://github.com/influxdata/telegraf/tree/release-1.6/plugins/inputs/prometheus to collect those metrics and push them to InfluxDB using https://github.com/influxdata/telegraf/tree/release-1.6/plugins/outputs/influxdb?

[fntlnz]

@ElvinEfendi that's a very different use case, the /metrics endpoint exposes collected and aggregated metrics related to the state of the ingress and the requests.
This on the other hand for every request sends a payload via udp to an InfluxDB telling some information about the request.

Worth noticing that you can also use telegraf instead of influxdb deployed as sidecar or at any other boundary if you don't want to send data to influxdb directly or you want to send the requests in batch of 10s 100s 1000s out of the node/network segment.

So the main differences are:

• This doesn't do any aggregation on the host, just an udp call
• This pushes the specific request infromation instead of pulling a summary of multiple requests

In the usecase we built it for it does not replace the /metrics endpoint, it is complimentary to it by giving access to the specific point in time which is the request.

[fntlnz]

@aledbf pushed the e2e tests, this is now ready for review.

[aledbf]

/approve
/lgtm

[aledbf]

@fntlnz thanks!

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: aledbf, fntlnz

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [aledbf]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ElvinEfendi]

@fntlnz thanks for the detailed use-case explanation - that makes sense!

[kfox1111]

Should the annotation be called something other then influxdb if that is the main use case? I could have easily overlooked it as I don't use influxdb.

[fntlnz]

@kfox1111 I don't understand what do you mean, can you give some examples?

[kfox1111]

@fntlnz the annotations are:
+|nginx.ingress.kubernetes.io/enable-influxdb|"true" or "false"|
+|nginx.ingress.kubernetes.io/influxdb-measurement|string|
+|nginx.ingress.kubernetes.io/influxdb-port|string|
+|nginx.ingress.kubernetes.io/influxdb-host|string|
+|nginx.ingress.kubernetes.io/influxdb-server-name|string|

But if you were not interested in using influxdb, but interested in "using udp to gather request statistics", its not obvious that you should be using the influx settings for that.

[aledbf]

@kfox1111 those settings ARE for influxdb. There is no "generic" nginx module for using UDP to gather request statistics, you need an specific module for that

[kfox1111]

But in an earlier comment, you said you could send it to telegraph instead of influxdb. so its not really an influxdb thing specifically?

[aledbf]

But in an earlier comment, you said you could send it to telegraph instead of influxdb. so its not really an influxdb thing specifically?

Yes, you can do that through an influxdb output plugin but this is done in influxdb, not nginx itself