Merge branch 'ci-gitlab-trace-metrics-errors' into 'master'

Rename metrics label for trace error metrics

See merge request gitlab-org/gitlab!72508

app/services/ci/update_build_state_service.rb

@@ -73,11 +73,11 @@ module Ci
       ::Gitlab::Ci::Trace::Checksum.new(build).then do |checksum|
         unless checksum.valid?
           metrics.increment_trace_operation(operation: :invalid)
-          metrics.increment_error_counter(type: :chunks_invalid_checksum)
+          metrics.increment_error_counter(error_reason: :chunks_invalid_checksum)
 
           if checksum.corrupted?
             metrics.increment_trace_operation(operation: :corrupted)
-            metrics.increment_error_counter(type: :chunks_invalid_size)
+            metrics.increment_error_counter(error_reason: :chunks_invalid_size)
           end
 
           next unless log_invalid_chunks?

doc/administration/monitoring/prometheus/gitlab_metrics.md

@@ -143,7 +143,7 @@ The following metrics are available:
 | `gitlab_snowplow_events_total`                                   | Counter     | 14.1    | Total number of GitLab Snowplow product intelligence events emitted                                                   |                                                           |
 | `gitlab_snowplow_failed_events_total`                            | Counter     | 14.1    | Total number of GitLab Snowplow product intelligence events emission failures                                         |                                                           |
 | `gitlab_snowplow_successful_events_total`                        | Counter     | 14.1    | Total number of GitLab Snowplow product intelligence events emission successes                                        |                                                           |
-| `gitlab_ci_build_trace_errors_total`                             | Counter     | 14.4    | Total amount of different error types on a build trace                                                                | `type`                                                    |
+| `gitlab_ci_build_trace_errors_total`                             | Counter     | 14.4    | Total amount of different error types on a build trace                                                                | `error_reason`                                            |
 
 ## Metrics controlled by a feature flag
 

lib/gitlab/ci/trace/archive.rb

@@ -62,7 +62,7 @@ module Gitlab
           trace_metadata.update!(remote_checksum: remote_checksum)
 
           unless trace_metadata.remote_checksum_valid?
-            metrics.increment_error_counter(type: :archive_invalid_checksum)
+            metrics.increment_error_counter(error_reason: :archive_invalid_checksum)
           end
         end
 

lib/gitlab/ci/trace/metrics.rb

@@ -21,7 +21,7 @@ module Gitlab
           :corrupted  # malformed trace found after comparing CRC32 and size
         ].freeze
 
-        TRACE_ERROR_TYPES = [
+        TRACE_ERROR_REASONS = [
           :chunks_invalid_size,       # used to be :corrupted
           :chunks_invalid_checksum,   # used to be :invalid
           :archive_invalid_checksum   # malformed trace found into object store after comparing MD5
@@ -39,12 +39,12 @@ module Gitlab
           self.class.trace_bytes.increment({}, size.to_i)
         end
 
-        def increment_error_counter(type: :unknown)
-          unless TRACE_ERROR_TYPES.include?(type)
-            raise ArgumentError, "unknown error type: #{type}"
+        def increment_error_counter(error_reason: :unknown)
+          unless TRACE_ERROR_REASONS.include?(error_reason)
+            raise ArgumentError, "unknown error reason: #{error_reason}"
           end
 
-          self.class.trace_errors_counter.increment(type: type)
+          self.class.trace_errors_counter.increment(error_reason: error_reason)
         end
 
         def observe_migration_duration(seconds)

spec/lib/gitlab/ci/trace/archive_spec.rb

@@ -40,7 +40,7 @@ RSpec.describe Gitlab::Ci::Trace::Archive do
           expect(trace_metadata.remote_checksum).to be_nil
           expect(metrics)
             .not_to have_received(:increment_error_counter)
-            .with(type: :archive_invalid_checksum)
+            .with(error_reason: :archive_invalid_checksum)
         end
       end
 
@@ -56,7 +56,7 @@ RSpec.describe Gitlab::Ci::Trace::Archive do
           expect(trace_metadata.remote_checksum).to be_nil
           expect(metrics)
             .not_to have_received(:increment_error_counter)
-            .with(type: :archive_invalid_checksum)
+            .with(error_reason: :archive_invalid_checksum)
         end
       end
 
@@ -72,7 +72,7 @@ RSpec.describe Gitlab::Ci::Trace::Archive do
           expect(trace_metadata.remote_checksum).to eq(src_checksum)
           expect(metrics)
             .not_to have_received(:increment_error_counter)
-            .with(type: :archive_invalid_checksum)
+            .with(error_reason: :archive_invalid_checksum)
         end
 
         context 'when the checksum does not match' do
@@ -92,7 +92,7 @@ RSpec.describe Gitlab::Ci::Trace::Archive do
             expect(trace_metadata.remote_checksum).to eq(invalid_remote_checksum)
             expect(metrics)
               .to have_received(:increment_error_counter)
-              .with(type: :archive_invalid_checksum)
+              .with(error_reason: :archive_invalid_checksum)
           end
         end
       end

spec/lib/gitlab/ci/trace/metrics_spec.rb

@@ -17,23 +17,23 @@ RSpec.describe Gitlab::Ci::Trace::Metrics, :prometheus do
   end
 
   describe '#increment_error_counter' do
-    context 'when the operation type is known' do
+    context 'when the error reason is known' do
       it 'increments the counter' do
-        subject.increment_error_counter(type: :chunks_invalid_size)
-        subject.increment_error_counter(type: :chunks_invalid_checksum)
-        subject.increment_error_counter(type: :archive_invalid_checksum)
+        subject.increment_error_counter(error_reason: :chunks_invalid_size)
+        subject.increment_error_counter(error_reason: :chunks_invalid_checksum)
+        subject.increment_error_counter(error_reason: :archive_invalid_checksum)
 
-        expect(described_class.trace_errors_counter.get(type: :chunks_invalid_size)).to eq 1
-        expect(described_class.trace_errors_counter.get(type: :chunks_invalid_checksum)).to eq 1
-        expect(described_class.trace_errors_counter.get(type: :archive_invalid_checksum)).to eq 1
+        expect(described_class.trace_errors_counter.get(error_reason: :chunks_invalid_size)).to eq 1
+        expect(described_class.trace_errors_counter.get(error_reason: :chunks_invalid_checksum)).to eq 1
+        expect(described_class.trace_errors_counter.get(error_reason: :archive_invalid_checksum)).to eq 1
 
         expect(described_class.trace_errors_counter.values.count).to eq 3
       end
     end
 
-    context 'when the operation type is known' do
+    context 'when the error reason is unknown' do
       it 'raises an exception' do
-        expect { subject.increment_error_counter(type: :invalid_type) }
+        expect { subject.increment_error_counter(error_reason: :invalid_type) }
           .to raise_error(ArgumentError)
       end
     end

spec/services/ci/update_build_state_service_spec.rb

@@ -118,7 +118,7 @@ RSpec.describe Ci::UpdateBuildStateService do
 
         expect(metrics)
           .not_to have_received(:increment_error_counter)
-          .with(type: :chunks_invalid_checksum)
+          .with(error_reason: :chunks_invalid_checksum)
       end
     end
 
@@ -188,7 +188,7 @@ RSpec.describe Ci::UpdateBuildStateService do
 
           expect(metrics)
             .to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_checksum)
+            .with(error_reason: :chunks_invalid_checksum)
         end
       end
 
@@ -210,11 +210,11 @@ RSpec.describe Ci::UpdateBuildStateService do
 
           expect(metrics)
             .not_to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_checksum)
+            .with(error_reason: :chunks_invalid_checksum)
 
           expect(metrics)
             .not_to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_size)
+            .with(error_reason: :chunks_invalid_size)
         end
 
         context 'when using deprecated parameters' do
@@ -235,11 +235,11 @@ RSpec.describe Ci::UpdateBuildStateService do
 
             expect(metrics)
               .not_to have_received(:increment_error_counter)
-              .with(type: :chunks_invalid_checksum)
+              .with(error_reason: :chunks_invalid_checksum)
 
             expect(metrics)
               .not_to have_received(:increment_error_counter)
-              .with(type: :chunks_invalid_size)
+              .with(error_reason: :chunks_invalid_size)
           end
         end
       end
@@ -262,11 +262,11 @@ RSpec.describe Ci::UpdateBuildStateService do
 
           expect(metrics)
             .to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_checksum)
+            .with(error_reason: :chunks_invalid_checksum)
 
           expect(metrics)
             .to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_size)
+            .with(error_reason: :chunks_invalid_size)
         end
       end
 
@@ -284,7 +284,7 @@ RSpec.describe Ci::UpdateBuildStateService do
 
           expect(metrics)
             .to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_checksum)
+            .with(error_reason: :chunks_invalid_checksum)
 
           expect(metrics)
             .not_to have_received(:increment_trace_operation)
@@ -292,7 +292,7 @@ RSpec.describe Ci::UpdateBuildStateService do
 
           expect(metrics)
             .not_to have_received(:increment_error_counter)
-            .with(type: :chunks_invalid_size)
+            .with(error_reason: :chunks_invalid_size)
         end
       end
 
@@ -376,7 +376,7 @@ RSpec.describe Ci::UpdateBuildStateService do
 
         expect(metrics)
           .not_to have_received(:increment_error_counter)
-          .with(type: :chunks_invalid_checksum)
+          .with(error_reason: :chunks_invalid_checksum)
       end
 
       context 'when build pending state is outdated' do