Configuring Cloudwatch Datasources

Grafana Configuration
1

Hello:

We are currently configuring the Cloudwatch data source.
Auth Provider: Access & secret key
Access key ID: Configured
Secret access key: Configured
Default Region: us-east-1
Custom Metrics: ?

What should go in the custom metrics sections if I want to be able to work with ebs, ec2, s3, efs ?

Thanks,

-Dan.

2

Hi,

You only need to fill in custom metrics section if you define and write custom metrics to certain namespaces. ebs, ec2, s3, efs should just work out of the box without any custom metrics. Give it a try and let me know if it works.

Marcus

3

Grafana
Grafana1913×879 51.3 KB

4

Hi,

Please remove everything from the Custom Metrics input field and then try Save & test.

Marcus

5

Grafana
Grafana1912×821 44.4 KB

6

Then it feels like there may be an permission issue on the AWS side. Do you find anything in the Grafana server logs when you get this error?

Are you running Grafana locally or a hosted Grafana instance?

Marcus

7

I’m running Zabbix, Postgres, Grafana on AWS m5.2xlarge ec2 instance. OS is RHEL7, FIPS compliant & selinux enforced. The credential I’m using are exactly the same one I’m using for aws cli commands. Below is an extract of the grafana logs:

t=2017-12-08T06:47:46-0600 lvl=eror msg=“Metric request error” logger=context userId=2 orgId=1 uname=username error="Failed to call cloudwatch:ListMetrics"
t=2017-12-08T06:47:46-0600 lvl=eror msg=“Request Completed” logger=context userId=2 orgId=1 uname=username method=POST path=/api/tsdb/query status=500 remote_addr=X.X.X.X time_ms=120307 size=34 referer=http://hostname.fqdn:3000/datasources/edit/2
t=2017-12-08T06:53:07-0600 lvl=eror msg=“Metric request error” logger=context userId=2 orgId=1 uname=username error="Failed to call cloudwatch:ListMetrics"
t=2017-12-08T06:53:07-0600 lvl=eror msg=“Request Completed” logger=context userId=2 orgId=1 uname=username method=POST path=/api/tsdb/query status=500 remote_addr=X.X.X.X time_ms=120375 size=34 referer=http://hostname.fqdn:3000/datasources/edit/2
t=2017-12-08T06:56:50-0600 lvl=eror msg=“Metric request error” logger=context userId=2 orgId=1 uname=username error="Failed to call cloudwatch:ListMetrics"
t=2017-12-08T06:56:50-0600 lvl=eror msg=“Request Completed” logger=context userId=2 orgId=1 uname=username method=POST path=/api/tsdb/query status=500 remote_addr=X.X.X.X time_ms=120355 size=34 referer=http://hostname.fqdn:3000/datasources/edit/2
t=2017-12-08T07:01:49-0600 lvl=eror msg=“Metric request error” logger=context userId=2 orgId=1 uname=username error="Failed to call cloudwatch:ListMetrics"
t=2017-12-08T07:01:49-0600 lvl=eror msg=“Request Completed” logger=context userId=2 orgId=1 uname=username method=POST path=/api/tsdb/query status=500 remote_addr=X.X.X.X time_ms=120427 size=34 referer=http://hostname.fqdn:3000/datasources/edit/2

8

Hi,

Sorry for late reply.

So I had a look at the issues and pull request regarding Cloudwatch datasource and I think that you’re problem is resolved in the master branch.

The following issue references the same error message that you have. Then there’s another credential issue. Both of these were merged to master 8 days ago by this pull request.

I suggest that you try out one of the latest nightly builds.

Marcus

9

Hello:

wget https://s3-us-west-2.amazonaws.com/grafana-releases/master/grafana-5.0.0-10925pre1.x86_64.rpm
sudo yum localinstall grafana-5.0.0-10925pre1.x86_64.rpm

Can I do an update to my grafana-4.6.3-1.x86_64 installed through yum repo.

10

Hi,

Please read the documentation here and here.

I would suggest that you backup your existing and/or doing a quick test using the Grafana docker image with the master tag.

Marcus