1

I have create a default alerting policy that notifies the user when the number of requests served by an external HTTP(S) load balancer is more than a certain threshold value.

In the Google cloud console, from the main navigation menu, I selected Monitoring > Alerting.

On the Alerting page, I clicked on Create policy.

In the Select a metric menu, the resource type for an external HTTP(S) load balancer that I have selected is https_lb_rule and the metric to monitor the number of requests served by the load balancer that I have selected is https/request_count.

For a trigger condition of type Threshold, I have entered the following information:
Alert trigger: Any time series violates
Threshold position: Above threshold
Threshold value: 1

Now I don't quite understand what does it mean to set the threshold value as 1. I was under the erroneous assumption that entering a value of 1 means that the alert is triggered if more than one request is served by the load balancer.

The email notification that I get for the incident says the following:

Request count for l7-http-gfev3 Google Cloud HTTP/S Load Balancing Rule labels {project_id=l7-http-gfev3, region=global, url_map_name=web-map-http, forwarding_rule_name=web-frontend-service, target_proxy_name=web-map-http-target-proxy, matched_url_path_rule=UNMATCHED, backend_target_name=web-backend-service} with metric labels {cache_result=MISS, client_country=United Kingdom, protocol=HTTP/1.1, proxy_continent=Europe, response_code=404, response_code_class=400} is above the threshold of 1.000 with a value of 5.000. The trailing segment of the message says: is above the threshold of 1.000 with a value of 5.000.

In another notification, the trailing segment of the message says: is above the threshold of 1.000 with a value of 35.000.

So, how should I interpret this message and what is the significance/meaning of threshold value of 1 or any other value in this context.

Nosail
  • 465
  • 2
  • 7
  • 19

1 Answers1

2

As you selected :

Alert trigger: Any time series violates 
Threshold position: Above threshold
Threshold value:1

Here, Threshold value indicates the cut off value. If you select above the threshold value then if there is any request more than 1 then the alert will generate. If you select below the threshold value then if there is a request less than 1 then the alert will generate. Any time if the series violates the threshold then the alert gets triggered.

As per your email, The trailing segment of the message says: is above the threshold of 1.000 with value of 5.000 which means request count is more than 1 and the request count at the time of email generation is 5.

In another notification, the trailing segment of the message says: is above the threshold of 1.000 with a value of 35.000 likewise above, request count is more than 1 and the count at the time of email generation the request count is 35.

Refer this link to know the Behavior of metric-based alerting policies

Edit 1 :

enter image description here

Hemanth Kumar
  • 2,728
  • 1
  • 4
  • 19
  • So, in this case, is it right to say that time-series data for `https/request_count` metric is synonymous with the request count? – Nosail Nov 15 '22 at 14:50
  • @Nosail : Yes it is a request count, you can see the attached screenshot under edit 1 in the answer section. if your external request is more than 1 (given threshold) then the alert will trigger to your mail box. – Hemanth Kumar Nov 16 '22 at 05:59