How to monitor HTTPS and TCP Load Balancers on GCP

google-cloud-platformgoogle-compute-enginegoogle-stackdriverload balancingmonitoring

we are using HTTPS and TCP Load balancers in the Google Cloud Platform (GCP) and we want to monitor them (e.g. with Stackdriver).

For us, one main criteria seems to be the amount of healthy backend services (which also is an inidicator shown in the GCP Console as green or orange/red).

But unfortunately there is no such metric as backend_service.count.

Is there any other metric or way which could be used for that kind of monitoring?
Is it overall a good idea to monitor and alert based on the backend count? (maybe we miss something)

Best Answer

I think what you’re looking for is not from the load-balancer itself, instead you may use Stackdriver Groups.

You may use in different ways the groups in SD, for example if you have an instance group, you may use tags in the nodes part of the instance group and create a group with all the resources in your project that contain the tag that you defined in your instance template.

Also, you may specify the membership criteria with multiple options, for example you may use multiple criterias in order to group the number of nodes and also the region where they are or any other option in the Create Group form.

In the SD monitoring console you will be able to see the existing nodes that match the criteria(s) being part of the group and a graphic with the Running Resources.

Related Topic