diff --git a/CHANGELOG.md b/CHANGELOG.md
index 6bea7065fa3..1e3f954b8db 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -49,6 +49,7 @@
   - `MimirContinuousTestNotRunningOnWrites`
   - `MimirContinuousTestNotRunningOnReads`
   - `MimirContinuousTestFailed`
+* [ENHANCEMENT] Added `per_cluster_label` support to allow to change the label name used to differentiate between Kubernetes clusters. #1651
 * [BUGFIX] Dashboards: Fix "Failed evaluation rate" panel on Tenants dashboard. #1629
 
 ### Jsonnet
diff --git a/docs/sources/operators-guide/visualizing-metrics/requirements.md b/docs/sources/operators-guide/visualizing-metrics/requirements.md
index dbe82e44323..3eb9def1920 100644
--- a/docs/sources/operators-guide/visualizing-metrics/requirements.md
+++ b/docs/sources/operators-guide/visualizing-metrics/requirements.md
@@ -13,7 +13,7 @@ The following table shows the required label names and whether they can be custo
 
 | Label name  | Configurable | Description                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      |
 | :---------- | :----------- | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `cluster`   | No           | The Kubernetes cluster or datacenter where the Mimir cluster is running.                                                                                                                                                                                                                                                                                                                                                                                                                         |
+| `cluster`   | Yes          | The Kubernetes cluster or datacenter where the Mimir cluster is running. The cluster label can be configured with the `per_cluster_label` field in the mixin config.                                                                                                                                                                                                                                                                                                                             |
 | `namespace` | No           | The Kubernetes namespace where the Mimir cluster is running.                                                                                                                                                                                                                                                                                                                                                                                                                                     |
 | `job`       | Partially    | The Kubernetes namespace and Mimir component in the format `<namespace>/<component>`. When running in monolithic mode, the `<component>` should be `mimir`. When running in microservices mode, the `<component>` should be the name of the specific Mimir component (singular), like `distributor`, `ingester` or `store-gateway`. The label name can't be configured, while the regular expressions used to match components can be configured with the `job_names` field in the mixin config. |
 | `pod`       | Yes          | The unique identifier of a Mimir replica (eg. Pod ID when running on Kubernetes). The label name can be configured with the `per_instance_label` field in the mixin config.                                                                                                                                                                                                                                                                                                                      |
diff --git a/operations/mimir-mixin/alerts/alerts.libsonnet b/operations/mimir-mixin/alerts/alerts.libsonnet
index 27bfd013097..db5fe341f61 100644
--- a/operations/mimir-mixin/alerts/alerts.libsonnet
+++ b/operations/mimir-mixin/alerts/alerts.libsonnet
@@ -435,7 +435,7 @@
           alert: $.alertName('ProvisioningTooManyWrites'),
           // 80k writes / s per ingester max.
           expr: |||
-            avg by (%(alert_aggregation_labels)s) (cluster_namespace_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m) > 80e3
+            avg by (%(alert_aggregation_labels)s) (%(alert_aggregation_rule_prefix)s_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m) > 80e3
           ||| % $._config,
           'for': '15m',
           labels: {
diff --git a/operations/mimir-mixin/alerts/blocks.libsonnet b/operations/mimir-mixin/alerts/blocks.libsonnet
index e54776cda4c..1feba86fdae 100644
--- a/operations/mimir-mixin/alerts/blocks.libsonnet
+++ b/operations/mimir-mixin/alerts/blocks.libsonnet
@@ -14,13 +14,13 @@
             (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (thanos_objstore_bucket_last_successful_upload_time{job=~".+/ingester.*"}) > 0)
             and
             # Only if the ingester has ingested samples over the last 4h.
-            (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (max_over_time(cluster_namespace_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m[4h])) > 0)
+            (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (max_over_time(%(alert_aggregation_rule_prefix)s_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m[4h])) > 0)
             and
             # Only if the ingester was ingesting samples 4h ago. This protects from the case the ingester instance
             # had ingested samples in the past, then no traffic was received for a long period and then it starts
             # receiving samples again. Without this check, the alert would fire as soon as it gets back receiving
             # samples, while the a block shipping is expected within the next 4h.
-            (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (max_over_time(cluster_namespace_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m[1h] offset 4h)) > 0)
+            (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (max_over_time(%(alert_aggregation_rule_prefix)s_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m[1h] offset 4h)) > 0)
           ||| % $._config,
           labels: {
             severity: 'critical',
@@ -37,7 +37,7 @@
           expr: |||
             (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (thanos_objstore_bucket_last_successful_upload_time{job=~".+/ingester.*"}) == 0)
             and
-            (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (max_over_time(cluster_namespace_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m[4h])) > 0)
+            (max by(%(alert_aggregation_labels)s, %(per_instance_label)s) (max_over_time(%(alert_aggregation_rule_prefix)s_%(per_instance_label)s:cortex_ingester_ingested_samples_total:rate1m[4h])) > 0)
           ||| % $._config,
           labels: {
             severity: 'critical',
diff --git a/operations/mimir-mixin/config.libsonnet b/operations/mimir-mixin/config.libsonnet
index 936bd0f67c1..f212383cacd 100644
--- a/operations/mimir-mixin/config.libsonnet
+++ b/operations/mimir-mixin/config.libsonnet
@@ -35,9 +35,12 @@
       overrides_exporter: 'overrides-exporter',
     },
 
+    // The label used to differentiate between different Kubernetes clusters.
+    per_cluster_label: 'cluster',
+
     // Grouping labels, to uniquely identify and group by {jobs, clusters}
-    job_labels: ['cluster', 'namespace', 'job'],
-    cluster_labels: ['cluster', 'namespace'],
+    job_labels: [$._config.per_cluster_label, 'namespace', 'job'],
+    cluster_labels: [$._config.per_cluster_label, 'namespace'],
 
     cortex_p99_latency_threshold_seconds: 2.5,
 
diff --git a/operations/mimir-mixin/dashboards/alertmanager.libsonnet b/operations/mimir-mixin/dashboards/alertmanager.libsonnet
index 015b9ef437d..e65c15efecd 100644
--- a/operations/mimir-mixin/dashboards/alertmanager.libsonnet
+++ b/operations/mimir-mixin/dashboards/alertmanager.libsonnet
@@ -11,11 +11,11 @@ local utils = import 'mixin-utils/utils.libsonnet';
        })
       .addPanel(
         $.panel('Total alerts') +
-        $.statPanel('sum(cluster_job_%s:cortex_alertmanager_alerts:sum{%s})' % [$._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)], format='short')
+        $.statPanel('sum(%s_job_%s:cortex_alertmanager_alerts:sum{%s})' % [$._config.per_cluster_label, $._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)], format='short')
       )
       .addPanel(
         $.panel('Total silences') +
-        $.statPanel('sum(cluster_job_%s:cortex_alertmanager_silences:sum{%s})' % [$._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)], format='short')
+        $.statPanel('sum(%s_job_%s:cortex_alertmanager_silences:sum{%s})' % [$._config.per_cluster_label, $._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)], format='short')
       )
       .addPanel(
         $.panel('Tenants') +
@@ -29,11 +29,11 @@ local utils = import 'mixin-utils/utils.libsonnet';
         $.queryPanel(
           [
             |||
-              sum(cluster_job:cortex_alertmanager_alerts_received_total:rate5m{%s})
+              sum(%s_job:cortex_alertmanager_alerts_received_total:rate5m{%s})
               -
-              sum(cluster_job:cortex_alertmanager_alerts_invalid_total:rate5m{%s})
-            ||| % [$.jobMatcher($._config.job_names.alertmanager), $.jobMatcher($._config.job_names.alertmanager)],
-            'sum(cluster_job:cortex_alertmanager_alerts_invalid_total:rate5m{%s})' % $.jobMatcher($._config.job_names.alertmanager),
+              sum(%s_job:cortex_alertmanager_alerts_invalid_total:rate5m{%s})
+            ||| % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager), $._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
+            'sum(%s_job:cortex_alertmanager_alerts_invalid_total:rate5m{%s})' % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
           ],
           ['success', 'failed']
         )
@@ -46,11 +46,11 @@ local utils = import 'mixin-utils/utils.libsonnet';
         $.queryPanel(
           [
             |||
-              sum(cluster_job_integration:cortex_alertmanager_notifications_total:rate5m{%s})
+              sum(%s_job_integration:cortex_alertmanager_notifications_total:rate5m{%s})
               -
-              sum(cluster_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s})
-            ||| % [$.jobMatcher($._config.job_names.alertmanager), $.jobMatcher($._config.job_names.alertmanager)],
-            'sum(cluster_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s})' % $.jobMatcher($._config.job_names.alertmanager),
+              sum(%s_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s})
+            ||| % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager), $._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
+            'sum(%s_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s})' % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
           ],
           ['success', 'failed']
         )
@@ -61,13 +61,13 @@ local utils = import 'mixin-utils/utils.libsonnet';
           [
             |||
               (
-              sum(cluster_job_integration:cortex_alertmanager_notifications_total:rate5m{%s}) by(integration)
+              sum(%s_job_integration:cortex_alertmanager_notifications_total:rate5m{%s}) by(integration)
               -
-              sum(cluster_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s}) by(integration)
+              sum(%s_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s}) by(integration)
               ) > 0
               or on () vector(0)
-            ||| % [$.jobMatcher($._config.job_names.alertmanager), $.jobMatcher($._config.job_names.alertmanager)],
-            'sum(cluster_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s}) by(integration)' % $.jobMatcher($._config.job_names.alertmanager),
+            ||| % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager), $._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
+            'sum(%s_job_integration:cortex_alertmanager_notifications_failed_total:rate5m{%s}) by(integration)' % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
           ],
           ['success - {{ integration }}', 'failed - {{ integration }}']
         )
@@ -104,7 +104,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
       .addPanel(
         $.panel('Per %s alerts' % $._config.per_instance_label) +
         $.queryPanel(
-          'sum by(%s) (cluster_job_%s:cortex_alertmanager_alerts:sum{%s})' % [$._config.per_instance_label, $._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)],
+          'sum by(%s) (%s_job_%s:cortex_alertmanager_alerts:sum{%s})' % [$._config.per_instance_label, $._config.per_cluster_label, $._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)],
           '{{%s}}' % $._config.per_instance_label
         ) +
         $.stack
@@ -112,7 +112,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
       .addPanel(
         $.panel('Per %s silences' % $._config.per_instance_label) +
         $.queryPanel(
-          'sum by(%s) (cluster_job_%s:cortex_alertmanager_silences:sum{%s})' % [$._config.per_instance_label, $._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)],
+          'sum by(%s) (%s_job_%s:cortex_alertmanager_silences:sum{%s})' % [$._config.per_instance_label, $._config.per_cluster_label, $._config.per_instance_label, $.jobMatcher($._config.job_names.alertmanager)],
           '{{%s}}' % $._config.per_instance_label
         ) +
         $.stack
@@ -205,11 +205,11 @@ local utils = import 'mixin-utils/utils.libsonnet';
         $.queryPanel(
           [
             |||
-              sum(cluster_job:cortex_alertmanager_state_replication_total:rate5m{%s})
+              sum(%s_job:cortex_alertmanager_state_replication_total:rate5m{%s})
               -
-              sum(cluster_job:cortex_alertmanager_state_replication_failed_total:rate5m{%s})
-            ||| % [$.jobMatcher($._config.job_names.alertmanager), $.jobMatcher($._config.job_names.alertmanager)],
-            'sum(cluster_job:cortex_alertmanager_state_replication_failed_total:rate5m{%s})' % $.jobMatcher($._config.job_names.alertmanager),
+              sum(%s_job:cortex_alertmanager_state_replication_failed_total:rate5m{%s})
+            ||| % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager), $._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
+            'sum(%s_job:cortex_alertmanager_state_replication_failed_total:rate5m{%s})' % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
           ],
           ['success', 'failed']
         )
@@ -219,11 +219,11 @@ local utils = import 'mixin-utils/utils.libsonnet';
         $.queryPanel(
           [
             |||
-              sum(cluster_job:cortex_alertmanager_partial_state_merges_total:rate5m{%s})
+              sum(%s_job:cortex_alertmanager_partial_state_merges_total:rate5m{%s})
               -
-              sum(cluster_job:cortex_alertmanager_partial_state_merges_failed_total:rate5m{%s})
-            ||| % [$.jobMatcher($._config.job_names.alertmanager), $.jobMatcher($._config.job_names.alertmanager)],
-            'sum(cluster_job:cortex_alertmanager_partial_state_merges_failed_total:rate5m{%s})' % $.jobMatcher($._config.job_names.alertmanager),
+              sum(%s_job:cortex_alertmanager_partial_state_merges_failed_total:rate5m{%s})
+            ||| % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager), $._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
+            'sum(%s_job:cortex_alertmanager_partial_state_merges_failed_total:rate5m{%s})' % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.alertmanager)],
           ],
           ['success', 'failed']
         )
diff --git a/operations/mimir-mixin/dashboards/dashboard-utils.libsonnet b/operations/mimir-mixin/dashboards/dashboard-utils.libsonnet
index 3fdcc72baf2..680e0eb5fee 100644
--- a/operations/mimir-mixin/dashboards/dashboard-utils.libsonnet
+++ b/operations/mimir-mixin/dashboards/dashboard-utils.libsonnet
@@ -54,17 +54,17 @@ local utils = import 'mixin-utils/utils.libsonnet';
           if $._config.singleBinary
           then d.addMultiTemplate('job', 'cortex_build_info', 'job')
           else d
-               .addMultiTemplate('cluster', 'cortex_build_info', 'cluster')
-               .addMultiTemplate('namespace', 'cortex_build_info{cluster=~"$cluster"}', 'namespace')
+               .addMultiTemplate('cluster', 'cortex_build_info', '%s' % $._config.per_cluster_label)
+               .addMultiTemplate('namespace', 'cortex_build_info{%s=~"$cluster"}' % $._config.per_cluster_label, 'namespace')
         else
           if $._config.singleBinary
           then d.addTemplate('job', 'cortex_build_info', 'job')
           else d
-               .addTemplate('cluster', 'cortex_build_info', 'cluster')
-               .addTemplate('namespace', 'cortex_build_info{cluster=~"$cluster"}', 'namespace'),
+               .addTemplate('cluster', 'cortex_build_info', '%s' % $._config.per_cluster_label)
+               .addTemplate('namespace', 'cortex_build_info{%s=~"$cluster"}' % $._config.per_cluster_label, 'namespace'),
 
       addActiveUserSelectorTemplates()::
-        self.addTemplate('user', 'cortex_ingester_active_series{cluster=~"$cluster", namespace=~"$namespace"}', 'user'),
+        self.addTemplate('user', 'cortex_ingester_active_series{%s=~"$cluster", namespace=~"$namespace"}' % $._config.per_cluster_label, 'user'),
 
       addCustomTemplate(name, values, defaultIndex=0):: self {
         templating+: {
@@ -99,17 +99,17 @@ local utils = import 'mixin-utils/utils.libsonnet';
   jobMatcher(job)::
     if $._config.singleBinary
     then 'job=~"$job"'
-    else 'cluster=~"$cluster", job=~"($namespace)/(%s)"' % job,
+    else '%s=~"$cluster", job=~"($namespace)/(%s)"' % [$._config.per_cluster_label, job],
 
   namespaceMatcher()::
     if $._config.singleBinary
     then 'job=~"$job"'
-    else 'cluster=~"$cluster", namespace=~"$namespace"',
+    else '%s=~"$cluster", namespace=~"$namespace"' % $._config.per_cluster_label,
 
   jobSelector(job)::
     if $._config.singleBinary
-    then [utils.selector.noop('cluster'), utils.selector.re('job', '$job')]
-    else [utils.selector.re('cluster', '$cluster'), utils.selector.re('job', '($namespace)/(%s)' % job)],
+    then [utils.selector.noop('%s' % $._config.per_cluster_label), utils.selector.re('job', '$job')]
+    else [utils.selector.re('%s' % $._config.per_cluster_label, '$cluster'), utils.selector.re('job', '($namespace)/(%s)' % job)],
 
   queryPanel(queries, legends, legendLink=null)::
     super.queryPanel(queries, legends, legendLink) + {
diff --git a/operations/mimir-mixin/dashboards/overrides.libsonnet b/operations/mimir-mixin/dashboards/overrides.libsonnet
index 29fb3749812..980ed068e38 100644
--- a/operations/mimir-mixin/dashboards/overrides.libsonnet
+++ b/operations/mimir-mixin/dashboards/overrides.libsonnet
@@ -13,7 +13,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
           datasource: '${datasource}',
           targets: [
             {
-              expr: 'max by(limit_name) (cortex_limits_defaults{cluster=~"$cluster",namespace=~"$namespace"})',
+              expr: 'max by(limit_name) (cortex_limits_defaults{%s=~"$cluster",namespace=~"$namespace"})' % $._config.per_cluster_label,
               instant: true,
               legendFormat: '',
               refId: 'A',
@@ -69,7 +69,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
           datasource: '${datasource}',
           targets: [
             {
-              expr: 'max by(user, limit_name) (cortex_limits_overrides{cluster=~"$cluster",namespace=~"$namespace",user=~"${tenant_id}"})',
+              expr: 'max by(user, limit_name) (cortex_limits_overrides{%s=~"$cluster",namespace=~"$namespace",user=~"${tenant_id}"})' % $._config.per_cluster_label,
               instant: true,
               legendFormat: '',
               refId: 'A',
diff --git a/operations/mimir-mixin/dashboards/rollout-progress.libsonnet b/operations/mimir-mixin/dashboards/rollout-progress.libsonnet
index fcefd096f38..d5c57a3c056 100644
--- a/operations/mimir-mixin/dashboards/rollout-progress.libsonnet
+++ b/operations/mimir-mixin/dashboards/rollout-progress.libsonnet
@@ -3,6 +3,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
 (import 'dashboard-utils.libsonnet') {
   local config = {
     namespace_matcher: $.namespaceMatcher(),
+    per_cluster_label: $._config.per_cluster_label,
     gateway_job_matcher: $.jobMatcher($._config.job_names.gateway),
     gateway_write_routes_regex: 'api_(v1|prom)_push',
     gateway_read_routes_regex: '(prometheus|api_prom)_api_v1_.+',
@@ -127,7 +128,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
 
         $.panel('Writes 99th latency') +
         $.newStatPanel(|||
-          histogram_quantile(0.99, sum by (le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_write_routes_regex)s"}))
+          histogram_quantile(0.99, sum by (le) (%(per_cluster_label)s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_write_routes_regex)s"}))
         ||| % config, unit='s', thresholds=[
           { color: 'green', value: null },
           { color: 'orange', value: 0.2 },
@@ -178,7 +179,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
 
         $.panel('Reads 99th latency') +
         $.newStatPanel(|||
-          histogram_quantile(0.99, sum by (le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_read_routes_regex)s"}))
+          histogram_quantile(0.99, sum by (le) (%(per_cluster_label)s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_read_routes_regex)s"}))
         ||| % config, unit='s', thresholds=[
           { color: 'green', value: null },
           { color: 'orange', value: 1 },
@@ -283,15 +284,15 @@ local utils = import 'mixin-utils/utils.libsonnet';
         $.panel('Latency vs 24h ago') +
         $.queryPanel([|||
           1 - (
-            avg_over_time(histogram_quantile(0.99, sum by (le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_write_routes_regex)s"} offset 24h))[1h:])
+            avg_over_time(histogram_quantile(0.99, sum by (le) (%(per_cluster_label)s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_write_routes_regex)s"} offset 24h))[1h:])
             /
-            avg_over_time(histogram_quantile(0.99, sum by (le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_write_routes_regex)s"}))[1h:])
+            avg_over_time(histogram_quantile(0.99, sum by (le) (%(per_cluster_label)s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_write_routes_regex)s"}))[1h:])
           )
         ||| % config, |||
           1 - (
-            avg_over_time(histogram_quantile(0.99, sum by (le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_read_routes_regex)s"} offset 24h))[1h:])
+            avg_over_time(histogram_quantile(0.99, sum by (le) (%(per_cluster_label)s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_read_routes_regex)s"} offset 24h))[1h:])
             /
-            avg_over_time(histogram_quantile(0.99, sum by (le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_read_routes_regex)s"}))[1h:])
+            avg_over_time(histogram_quantile(0.99, sum by (le) (%(per_cluster_label)s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%(gateway_job_matcher)s, route=~"%(gateway_read_routes_regex)s"}))[1h:])
           )
         ||| % config], ['writes', 'reads']) + {
           yaxes: $.yaxes({
diff --git a/operations/mimir-mixin/dashboards/ruler.libsonnet b/operations/mimir-mixin/dashboards/ruler.libsonnet
index 6206bfe64a1..7bf6bb292a5 100644
--- a/operations/mimir-mixin/dashboards/ruler.libsonnet
+++ b/operations/mimir-mixin/dashboards/ruler.libsonnet
@@ -116,7 +116,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
       .addPanel(
         $.panel('Per route p99 latency') +
         $.queryPanel(
-          'histogram_quantile(0.99, sum by (route, le) (cluster_job_route:cortex_request_duration_seconds_bucket:sum_rate{%s, route=~"%s"}))' % [$.jobMatcher($._config.job_names.gateway), ruler_config_api_routes_re],
+          'histogram_quantile(0.99, sum by (route, le) (%s_job_route:cortex_request_duration_seconds_bucket:sum_rate{%s, route=~"%s"}))' % [$._config.per_cluster_label, $.jobMatcher($._config.job_names.gateway), ruler_config_api_routes_re],
           '{{ route }}'
         ) +
         { yaxes: $.yaxes('s') }
diff --git a/operations/mimir-mixin/dashboards/scaling.libsonnet b/operations/mimir-mixin/dashboards/scaling.libsonnet
index 36dead45cf6..67980d16ff0 100644
--- a/operations/mimir-mixin/dashboards/scaling.libsonnet
+++ b/operations/mimir-mixin/dashboards/scaling.libsonnet
@@ -41,11 +41,11 @@ local utils = import 'mixin-utils/utils.libsonnet';
         $.tablePanel([
           |||
             sort_desc(
-              cluster_namespace_deployment_reason:required_replicas:count{cluster=~"$cluster", namespace=~"$namespace"}
+              %s_deployment_reason:required_replicas:count{%s}
                 > ignoring(reason) group_left
-              cluster_namespace_deployment:actual_replicas:count{cluster=~"$cluster", namespace=~"$namespace"}
+              %s_deployment:actual_replicas:count{%s}
             )
-          |||,
+          ||| % [$._config.alert_aggregation_rule_prefix, $.namespaceMatcher(), $._config.alert_aggregation_rule_prefix, $.namespaceMatcher()],
         ], {
           __name__: { alias: 'Cluster', type: 'hidden' },
           cluster: { alias: 'Cluster' },
diff --git a/operations/mimir-mixin/dashboards/slow-queries.libsonnet b/operations/mimir-mixin/dashboards/slow-queries.libsonnet
index f28805b8960..f11cdd8f4d0 100644
--- a/operations/mimir-mixin/dashboards/slow-queries.libsonnet
+++ b/operations/mimir-mixin/dashboards/slow-queries.libsonnet
@@ -16,7 +16,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
           targets: [
             {
               // Filter out the remote read endpoint.
-              expr: '{cluster=~"$cluster",namespace=~"$namespace",name=~"query-frontend.*"} |= "query stats" != "/api/v1/read" | logfmt | org_id=~"${tenant_id}" | response_time > ${min_duration}',
+              expr: '{%s=~"$cluster",namespace=~"$namespace",name=~"query-frontend.*"} |= "query stats" != "/api/v1/read" | logfmt | org_id=~"${tenant_id}" | response_time > ${min_duration}' % $._config.per_cluster_label,
               instant: false,
               legendFormat: '',
               range: true,
diff --git a/operations/mimir-mixin/groups.libsonnet b/operations/mimir-mixin/groups.libsonnet
index c2c35f90d21..a88b7e4c42f 100644
--- a/operations/mimir-mixin/groups.libsonnet
+++ b/operations/mimir-mixin/groups.libsonnet
@@ -58,5 +58,14 @@
           ),
         ),
       ),
+    alert_aggregation_rule_prefix:
+      std.join(
+        '_',
+        // Split the configured labels by comma and remove whitespaces.
+        std.map(
+          function(l) std.strReplace(l, ' ', ''),
+          std.split($._config.alert_aggregation_labels, ',')
+        ),
+      ),
   },
 }
diff --git a/operations/mimir-mixin/recording_rules.libsonnet b/operations/mimir-mixin/recording_rules.libsonnet
index 40e8e6cf59d..b2878681447 100644
--- a/operations/mimir-mixin/recording_rules.libsonnet
+++ b/operations/mimir-mixin/recording_rules.libsonnet
@@ -12,45 +12,45 @@ local utils = import 'mixin-utils/utils.libsonnet';
       {
         name: 'mimir_api_1',
         rules:
-          utils.histogramRules('cortex_request_duration_seconds', ['cluster', 'job']),
+          utils.histogramRules('cortex_request_duration_seconds', [$._config.per_cluster_label, 'job']),
       },
       {
         name: 'mimir_api_2',
         rules:
-          utils.histogramRules('cortex_request_duration_seconds', ['cluster', 'job', 'route']),
+          utils.histogramRules('cortex_request_duration_seconds', [$._config.per_cluster_label, 'job', 'route']),
       },
       {
         name: 'mimir_api_3',
         rules:
-          utils.histogramRules('cortex_request_duration_seconds', ['cluster', 'namespace', 'job', 'route']),
+          utils.histogramRules('cortex_request_duration_seconds', $._config.job_labels + ['route']),
       },
       {
         name: 'mimir_querier_api',
         rules:
-          utils.histogramRules('cortex_querier_request_duration_seconds', ['cluster', 'job']) +
-          utils.histogramRules('cortex_querier_request_duration_seconds', ['cluster', 'job', 'route']) +
-          utils.histogramRules('cortex_querier_request_duration_seconds', ['cluster', 'namespace', 'job', 'route']),
+          utils.histogramRules('cortex_querier_request_duration_seconds', [$._config.per_cluster_label, 'job']) +
+          utils.histogramRules('cortex_querier_request_duration_seconds', [$._config.per_cluster_label, 'job', 'route']) +
+          utils.histogramRules('cortex_querier_request_duration_seconds', $._config.job_labels + ['route']),
       },
       {
         name: 'mimir_cache',
         rules:
-          utils.histogramRules('cortex_memcache_request_duration_seconds', ['cluster', 'job', 'method']) +
-          utils.histogramRules('cortex_cache_request_duration_seconds', ['cluster', 'job']) +
-          utils.histogramRules('cortex_cache_request_duration_seconds', ['cluster', 'job', 'method']),
+          utils.histogramRules('cortex_memcache_request_duration_seconds', [$._config.per_cluster_label, 'job', 'method']) +
+          utils.histogramRules('cortex_cache_request_duration_seconds', [$._config.per_cluster_label, 'job']) +
+          utils.histogramRules('cortex_cache_request_duration_seconds', [$._config.per_cluster_label, 'job', 'method']),
       },
       {
         name: 'mimir_storage',
         rules:
-          utils.histogramRules('cortex_kv_request_duration_seconds', ['cluster', 'job']),
+          utils.histogramRules('cortex_kv_request_duration_seconds', [$._config.per_cluster_label, 'job']),
       },
       {
         name: 'mimir_queries',
         rules:
-          utils.histogramRules('cortex_query_frontend_retries', ['cluster', 'job']) +
-          utils.histogramRules('cortex_query_frontend_queue_duration_seconds', ['cluster', 'job']) +
-          utils.histogramRules('cortex_ingester_queried_series', ['cluster', 'job']) +
-          utils.histogramRules('cortex_ingester_queried_samples', ['cluster', 'job']) +
-          utils.histogramRules('cortex_ingester_queried_exemplars', ['cluster', 'job']),
+          utils.histogramRules('cortex_query_frontend_retries', [$._config.per_cluster_label, 'job']) +
+          utils.histogramRules('cortex_query_frontend_queue_duration_seconds', [$._config.per_cluster_label, 'job']) +
+          utils.histogramRules('cortex_ingester_queried_series', [$._config.per_cluster_label, 'job']) +
+          utils.histogramRules('cortex_ingester_queried_samples', [$._config.per_cluster_label, 'job']) +
+          utils.histogramRules('cortex_ingester_queried_exemplars', [$._config.per_cluster_label, 'job']),
       },
       {
         name: 'mimir_received_samples',
@@ -113,9 +113,9 @@ local utils = import 'mixin-utils/utils.libsonnet';
           {
             // Convenience rule to get the number of replicas for both a deployment and a statefulset.
             // Multi-zone deployments are grouped together removing the "zone-X" suffix.
-            record: 'cluster_namespace_deployment:actual_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment:actual_replicas:count' % _config,
             expr: |||
-              sum by (cluster, namespace, deployment) (
+              sum by (%(alert_aggregation_labels)s, deployment) (
                 label_replace(
                   kube_deployment_spec_replicas,
                   # The question mark in "(.*?)" is used to make it non-greedy, otherwise it
@@ -124,14 +124,14 @@ local utils = import 'mixin-utils/utils.libsonnet';
                 )
               )
               or
-              sum by (cluster, namespace, deployment) (
+              sum by (%(alert_aggregation_labels)s, deployment) (
                 label_replace(kube_statefulset_replicas, "deployment", "$1", "statefulset", "(.*?)(?:-zone-[a-z])?")
               )
-            |||,
+            ||| % _config,
           },
           {
             // Distributors should be able to deal with 240k samples/s.
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'distributor',
               reason: 'sample_rate',
@@ -139,7 +139,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
             expr: |||
               ceil(
                 quantile_over_time(0.99,
-                  sum by (cluster, namespace) (
+                  sum by (%(alert_aggregation_labels)s) (
                     %(group_prefix_jobs)s:cortex_distributor_received_samples:rate5m
                   )[24h:]
                 )
@@ -150,14 +150,14 @@ local utils = import 'mixin-utils/utils.libsonnet';
           {
             // We should be about to cover 80% of our limits,
             // and ingester can have 80k samples/s.
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'distributor',
               reason: 'sample_rate_limits',
             },
             expr: |||
               ceil(
-                sum by (cluster, namespace) (cortex_limits_overrides{limit_name="ingestion_rate"})
+                sum by (%(alert_aggregation_labels)s) (cortex_limits_overrides{limit_name="ingestion_rate"})
                 * %(limit_utilisation_target)s / %(max_samples_per_sec_per_distributor)s
               )
             ||| % _config,
@@ -165,7 +165,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
           {
             // We want ingesters each ingester to deal with 80k samples/s.
             // NB we measure this at the distributors and multiple by RF (3).
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'ingester',
               reason: 'sample_rate',
@@ -173,7 +173,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
             expr: |||
               ceil(
                 quantile_over_time(0.99,
-                  sum by (cluster, namespace) (
+                  sum by (%(alert_aggregation_labels)s) (
                     %(group_prefix_jobs)s:cortex_distributor_received_samples:rate5m
                   )[24h:]
                 )
@@ -183,7 +183,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
           },
           {
             // Ingester should have 1.5M series in memory
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'ingester',
               reason: 'active_series',
@@ -191,7 +191,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
             expr: |||
               ceil(
                 quantile_over_time(0.99,
-                  sum by(cluster, namespace) (
+                  sum by(%(alert_aggregation_labels)s) (
                     cortex_ingester_memory_series
                   )[24h:]
                 )
@@ -202,14 +202,14 @@ local utils = import 'mixin-utils/utils.libsonnet';
           {
             // We should be about to cover 60% of our limits,
             // and ingester can have 1.5M series in memory
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'ingester',
               reason: 'active_series_limits',
             },
             expr: |||
               ceil(
-                sum by (cluster, namespace) (cortex_limits_overrides{limit_name="max_global_series_per_user"})
+                sum by (%(alert_aggregation_labels)s) (cortex_limits_overrides{limit_name="max_global_series_per_user"})
                 * 3 * %(limit_utilisation_target)s / %(max_series_per_ingester)s
               )
             ||| % _config,
@@ -217,43 +217,43 @@ local utils = import 'mixin-utils/utils.libsonnet';
           {
             // We should be about to cover 60% of our limits,
             // and ingester can have 80k samples/s.
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'ingester',
               reason: 'sample_rate_limits',
             },
             expr: |||
               ceil(
-                sum by (cluster, namespace) (cortex_limits_overrides{limit_name="ingestion_rate"})
+                sum by (%(alert_aggregation_labels)s) (cortex_limits_overrides{limit_name="ingestion_rate"})
                 * %(limit_utilisation_target)s / %(max_samples_per_sec_per_ingester)s
               )
             ||| % _config,
           },
           {
             // Ingesters store 96h of data on disk - we want memcached to store 1/4 of that.
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               deployment: 'memcached',
               reason: 'active_series',
             },
             expr: |||
               ceil(
-                (sum by (cluster, namespace) (
+                (sum by (%(alert_aggregation_labels)s) (
                   cortex_ingester_tsdb_storage_blocks_bytes{job=~".+/ingester.*"}
                 ) / 4)
                   /
-                avg by (cluster, namespace) (
+                avg by (%(alert_aggregation_labels)s) (
                   memcached_limit_bytes{job=~".+/memcached"}
                 )
               )
-            |||,
+            ||| % _config,
           },
           {
             // Convenience rule to get the CPU utilization for both a deployment and a statefulset.
             // Multi-zone deployments are grouped together removing the "zone-X" suffix.
-            record: 'cluster_namespace_deployment:container_cpu_usage_seconds_total:sum_rate',
+            record: '%(alert_aggregation_rule_prefix)s_deployment:container_cpu_usage_seconds_total:sum_rate' % _config,
             expr: |||
-              sum by (cluster, namespace, deployment) (
+              sum by (%(alert_aggregation_labels)s, deployment) (
                 label_replace(
                   label_replace(
                     node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate,
@@ -264,12 +264,12 @@ local utils = import 'mixin-utils/utils.libsonnet';
                   "deployment", "$1", "deployment", "(.*?)(?:-zone-[a-z])?"
                 )
               )
-            |||,
+            ||| % _config,
           },
           {
             // Convenience rule to get the CPU request for both a deployment and a statefulset.
             // Multi-zone deployments are grouped together removing the "zone-X" suffix.
-            record: 'cluster_namespace_deployment:kube_pod_container_resource_requests_cpu_cores:sum',
+            record: '%(alert_aggregation_rule_prefix)s_deployment:kube_pod_container_resource_requests_cpu_cores:sum' % _config,
             expr: |||
               # This recording rule is made compatible with the breaking changes introduced in kube-state-metrics v2
               # that remove resource metrics, ref:
@@ -279,7 +279,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
               # This is the old expression, compatible with kube-state-metrics < v2.0.0,
               # where kube_pod_container_resource_requests_cpu_cores was removed:
               (
-                sum by (cluster, namespace, deployment) (
+                sum by (%(alert_aggregation_labels)s, deployment) (
                   label_replace(
                     label_replace(
                       kube_pod_container_resource_requests_cpu_cores,
@@ -295,7 +295,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
               # This expression is compatible with kube-state-metrics >= v1.4.0,
               # where kube_pod_container_resource_requests was introduced.
               (
-                sum by (cluster, namespace, deployment) (
+                sum by (%(alert_aggregation_labels)s, deployment) (
                   label_replace(
                     label_replace(
                       kube_pod_container_resource_requests{resource="cpu"},
@@ -307,32 +307,32 @@ local utils = import 'mixin-utils/utils.libsonnet';
                   )
                 )
               )
-            |||,
+            ||| % _config,
           },
           {
             // Jobs should be sized to their CPU usage.
             // We do this by comparing 99th percentile usage over the last 24hrs to
             // their current provisioned #replicas and resource requests.
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               reason: 'cpu_usage',
             },
             expr: |||
               ceil(
-                cluster_namespace_deployment:actual_replicas:count
+                %(alert_aggregation_rule_prefix)s_deployment:actual_replicas:count
                   *
-                quantile_over_time(0.99, cluster_namespace_deployment:container_cpu_usage_seconds_total:sum_rate[24h])
+                quantile_over_time(0.99, %(alert_aggregation_rule_prefix)s_deployment:container_cpu_usage_seconds_total:sum_rate[24h])
                   /
-                cluster_namespace_deployment:kube_pod_container_resource_requests_cpu_cores:sum
+                %(alert_aggregation_rule_prefix)s_deployment:kube_pod_container_resource_requests_cpu_cores:sum
               )
-            |||,
+            ||| % _config,
           },
           {
             // Convenience rule to get the Memory utilization for both a deployment and a statefulset.
             // Multi-zone deployments are grouped together removing the "zone-X" suffix.
-            record: 'cluster_namespace_deployment:container_memory_usage_bytes:sum',
+            record: '%(alert_aggregation_rule_prefix)s_deployment:container_memory_usage_bytes:sum' % _config,
             expr: |||
-              sum by (cluster, namespace, deployment) (
+              sum by (%(alert_aggregation_labels)s, deployment) (
                 label_replace(
                   label_replace(
                     container_memory_usage_bytes,
@@ -343,12 +343,12 @@ local utils = import 'mixin-utils/utils.libsonnet';
                   "deployment", "$1", "deployment", "(.*?)(?:-zone-[a-z])?"
                 )
               )
-            |||,
+            ||| % _config,
           },
           {
             // Convenience rule to get the Memory request for both a deployment and a statefulset.
             // Multi-zone deployments are grouped together removing the "zone-X" suffix.
-            record: 'cluster_namespace_deployment:kube_pod_container_resource_requests_memory_bytes:sum',
+            record: '%(alert_aggregation_rule_prefix)s_deployment:kube_pod_container_resource_requests_memory_bytes:sum' % _config,
             expr: |||
               # This recording rule is made compatible with the breaking changes introduced in kube-state-metrics v2
               # that remove resource metrics, ref:
@@ -358,7 +358,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
               # This is the old expression, compatible with kube-state-metrics < v2.0.0,
               # where kube_pod_container_resource_requests_memory_bytes was removed:
               (
-                sum by (cluster, namespace, deployment) (
+                sum by (%(alert_aggregation_labels)s, deployment) (
                   label_replace(
                     label_replace(
                       kube_pod_container_resource_requests_memory_bytes,
@@ -374,7 +374,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
               # This expression is compatible with kube-state-metrics >= v1.4.0,
               # where kube_pod_container_resource_requests was introduced.
               (
-                sum by (cluster, namespace, deployment) (
+                sum by (%(alert_aggregation_labels)s, deployment) (
                   label_replace(
                     label_replace(
                       kube_pod_container_resource_requests{resource="memory"},
@@ -386,25 +386,25 @@ local utils = import 'mixin-utils/utils.libsonnet';
                   )
                 )
               )
-            |||,
+            ||| % _config,
           },
           {
             // Jobs should be sized to their Memory usage.
             // We do this by comparing 99th percentile usage over the last 24hrs to
             // their current provisioned #replicas and resource requests.
-            record: 'cluster_namespace_deployment_reason:required_replicas:count',
+            record: '%(alert_aggregation_rule_prefix)s_deployment_reason:required_replicas:count' % _config,
             labels: {
               reason: 'memory_usage',
             },
             expr: |||
               ceil(
-                cluster_namespace_deployment:actual_replicas:count
+                %(alert_aggregation_rule_prefix)s_deployment:actual_replicas:count
                   *
-                quantile_over_time(0.99, cluster_namespace_deployment:container_memory_usage_bytes:sum[24h])
+                quantile_over_time(0.99, %(alert_aggregation_rule_prefix)s_deployment:container_memory_usage_bytes:sum[24h])
                   /
-                cluster_namespace_deployment:kube_pod_container_resource_requests_memory_bytes:sum
+                %(alert_aggregation_rule_prefix)s_deployment:kube_pod_container_resource_requests_memory_bytes:sum
               )
-            |||,
+            ||| % _config,
           },
         ],
       },
@@ -413,64 +413,64 @@ local utils = import 'mixin-utils/utils.libsonnet';
         rules: [
           // Aggregations of per-user Alertmanager metrics used in dashboards.
           {
-            record: 'cluster_job_%s:cortex_alertmanager_alerts:sum' % $._config.per_instance_label,
+            record: '%s_job_%s:cortex_alertmanager_alerts:sum' % [$._config.per_cluster_label, $._config.per_instance_label],
             expr: |||
-              sum by (cluster, job, %s) (cortex_alertmanager_alerts)
-            ||| % $._config.per_instance_label,
+              sum by (%s, job, %s) (cortex_alertmanager_alerts)
+            ||| % [$._config.per_cluster_label, $._config.per_instance_label],
           },
           {
-            record: 'cluster_job_%s:cortex_alertmanager_silences:sum' % $._config.per_instance_label,
+            record: '%s_job_%s:cortex_alertmanager_silences:sum' % [$._config.per_cluster_label, $._config.per_instance_label],
             expr: |||
-              sum by (cluster, job, %s) (cortex_alertmanager_silences)
-            ||| % $._config.per_instance_label,
+              sum by (%s, job, %s) (cortex_alertmanager_silences)
+            ||| % [$._config.per_cluster_label, $._config.per_instance_label],
           },
           {
-            record: 'cluster_job:cortex_alertmanager_alerts_received_total:rate5m',
+            record: '%s_job:cortex_alertmanager_alerts_received_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job) (rate(cortex_alertmanager_alerts_received_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job) (rate(cortex_alertmanager_alerts_received_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job:cortex_alertmanager_alerts_invalid_total:rate5m',
+            record: '%s_job:cortex_alertmanager_alerts_invalid_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job) (rate(cortex_alertmanager_alerts_invalid_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job) (rate(cortex_alertmanager_alerts_invalid_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job_integration:cortex_alertmanager_notifications_total:rate5m',
+            record: '%s_job_integration:cortex_alertmanager_notifications_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job, integration) (rate(cortex_alertmanager_notifications_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job, integration) (rate(cortex_alertmanager_notifications_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job_integration:cortex_alertmanager_notifications_failed_total:rate5m',
+            record: '%s_job_integration:cortex_alertmanager_notifications_failed_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job, integration) (rate(cortex_alertmanager_notifications_failed_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job, integration) (rate(cortex_alertmanager_notifications_failed_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job:cortex_alertmanager_state_replication_total:rate5m',
+            record: '%s_job:cortex_alertmanager_state_replication_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job) (rate(cortex_alertmanager_state_replication_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job) (rate(cortex_alertmanager_state_replication_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job:cortex_alertmanager_state_replication_failed_total:rate5m',
+            record: '%s_job:cortex_alertmanager_state_replication_failed_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job) (rate(cortex_alertmanager_state_replication_failed_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job) (rate(cortex_alertmanager_state_replication_failed_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job:cortex_alertmanager_partial_state_merges_total:rate5m',
+            record: '%s_job:cortex_alertmanager_partial_state_merges_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job) (rate(cortex_alertmanager_partial_state_merges_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job) (rate(cortex_alertmanager_partial_state_merges_total[5m]))
+            ||| % _config,
           },
           {
-            record: 'cluster_job:cortex_alertmanager_partial_state_merges_failed_total:rate5m',
+            record: '%s_job:cortex_alertmanager_partial_state_merges_failed_total:rate5m' % $._config.per_cluster_label,
             expr: |||
-              sum by (cluster, job) (rate(cortex_alertmanager_partial_state_merges_failed_total[5m]))
-            |||,
+              sum by (%(per_cluster_label)s, job) (rate(cortex_alertmanager_partial_state_merges_failed_total[5m]))
+            ||| % _config,
           },
         ],
       },
@@ -479,7 +479,7 @@ local utils = import 'mixin-utils/utils.libsonnet';
         rules: [
           {
             // cortex_ingester_ingested_samples_total is per user, in this rule we want to see the sum per cluster/namespace/instance
-            record: 'cluster_namespace_%s:cortex_ingester_ingested_samples_total:rate1m' % $._config.per_instance_label,
+            record: '%s_%s:cortex_ingester_ingested_samples_total:rate1m' % [$._config.alert_aggregation_rule_prefix, $._config.per_instance_label],
             expr: |||
               sum by(%(alert_aggregation_labels)s, %(per_instance_label)s) (rate(cortex_ingester_ingested_samples_total[1m]))
             ||| % $._config,