Adding sinks configurations

Changing playbooks output to sinks
Removed slack mentions
Add prometheus_url override global param
Fix node cpu analysis prometheus query

.gitignore

@@ -1,6 +1,7 @@
 .idea/
 .vscode/
 src/build/
+playground/
 src/dist/
 *.egg/
 *.egg-info/

docs/developer-guide/reference.rst

@@ -95,3 +95,47 @@ For example:
     @on_recurring_trigger(seconds_delay=10, repeat=3)
     def my_scheduled_playbook(event: RecurringTriggerEvent):
         logging.info(f"My scheduled playbook is running for the {event.recurrence} time")
+
+Sinks
+-------------
+| Playbooks results can be forwarded to one or more sinks. See :ref:`Playbooks sinks` for details.
+| For that to happen, we have to create ``Finding`` and ``Enrichments`` during the playbooks processing.
+| The Robusta platform will automatically forward it to the configured sinks
+
+Finding
+^^^^^^^^^^^^^^^^^
+| The ``Finding`` contains the base information of the playbooks result.
+| Creating a ``Finding`` is easy:
+
+.. code-block:: python
+
+    @on_recurring_trigger(seconds_delay=10, repeat=3)
+    def my_scheduled_playbook(event: RecurringTriggerEvent):
+        event.processing_context.create_finding(
+            title=f"My scheduled playbook is running for the {event.recurrence} time",
+            severity=FindingSeverity.INFO
+    )
+
+Enrichments
+^^^^^^^^^^^^^^^^^
+| Each ``Finding`` can contain any number of ``Enrichments``.
+| Each ``Enrichment`` has a list of ``blocks`` describing it:
+* **MarkdownBlock:** - A text block
+* **DividerBlock:** - Dividing section between ``Enrichment`` parts. (If the sink supports that)
+* **HeaderBlock:** - A header block
+* **ListBlock:** - A block containing list of items
+* **TableBlock:** - A block containing table of items
+* **KubernetesFieldsBlock:** - A block containing information describing kubernetes fields
+* **DiffsBlock:** - A block containing information describing yaml diff attributes
+* **JsonBlock:** - A block containing any json data
+* **FileBlock:** - A block containing any file with the file data
+* **CallbackBlock:** - A block containing callback information, that can be invoked back from the sink. (a Slack button for example, running some command)
+
+| **Note:** - Not all block types supported by all sinks. If an unsupported block arrives to a sink, it will be ignored
+
+| Adding an ``Enrichment``:
+
+.. code-block:: python
+
+    my_log_file_data = "..."
+    event.processing_context.finding.add_enrichment([FileBlock("log.txt", my_log_file_data)])

docs/getting-started/manual-triggers.rst

@@ -25,4 +25,4 @@ Manually triggering the python_profiler playbook
 
 .. code-block:: bash
 
-    robusta trigger python_profiler pod_name=<POD_NAME> namespace=<NAMESPACE> slack_channel=<SLACK_CHANNEL>
+    robusta trigger python_profiler pod_name=<POD_NAME> namespace=<NAMESPACE>

docs/user-guide/alerts.rst

@@ -35,8 +35,6 @@ Lets look at the simplest possible ``active_playbooks.yaml`` which instructs Rob
 
    active_playbooks:
   - name: "alerts_integration"
-    action_params:
-      slack_channel: "robusta-alerts"
 
 The above configuration isn't very useful because we haven't enriched any alerts yet.
 However, we do get a minor aesthetic benefit because Robusta adds pretty formatting to alerts as you can see below:
@@ -56,7 +54,6 @@ Now lets add an enricher to ``active_playbooks.yaml`` which enriches the ``HostH
    active_playbooks:
   - name: "alerts_integration"
     action_params:
-      slack_channel: "robusta-alerts"
       alerts_config:
       - alert_name: "HostHighCpuLoad"
         enrichers:
@@ -90,7 +87,6 @@ You can change the default enricher(s) for all alerts using the ``default_enrich
    active_playbooks:
   - name: "alerts_integration"
     action_params:
-      slack_channel: "robusta-alerts"
       default_enrichers:
         - name: "AlertDefaults"
 
@@ -105,7 +101,6 @@ Now lets look at an example ``active_playbooks.yaml`` which silences KubePodCras
    active_playbooks:
   - name: "alerts_integration"
     action_params:
-      slack_channel: "robusta-alerts"
       alerts_config:
       - alert_name: "KubePodCrashLooping"
         silencers:
@@ -122,7 +117,6 @@ Here is an example which shows all the features discussed above working together
    active_playbooks:
   - name: "alerts_integration"
     action_params:
-      slack_channel: "robusta-alerts"
       default_enrichers:
         - name: "AlertDefaults"
       alerts_config:
@@ -145,10 +139,14 @@ Available enrichers
 **AlertDefaults:** send the alert message and labels to Slack
 
 **NodeCPUAnalysis:** provide deep analysis of node cpu usage
+This enricher use ``prometheus``. The ``prometheus`` url can be overriden in the ``global_config`` section.
+(For example - ``prometheus_url: "http://prometheus-k8s.monitoring.svc.cluster.local:9090"``)
 
 **OOMKillerEnricher:** shows which pods were recently OOM Killed on a node
 
 **GraphEnricher:** display a graph of the Prometheus query which triggered the alert
+This enricher use ``prometheus``. The ``prometheus`` url can be overriden in the ``global_config`` section.
+(For example - ``prometheus_url: "http://prometheus-k8s.monitoring.svc.cluster.local:9090"``)
 
 **StackOverflowEnricher:** add a button in Slack to search for the alert name on StackOverflow
 

docs/user-guide/builtin-playbooks.rst

@@ -90,7 +90,7 @@ python_profiler
 
 .. code-block:: bash
 
-   robusta trigger python_profiler pod_name=your-pod namespace=you-ns process_name=your-process slack_channel=general seconds=5
+   robusta trigger python_profiler pod_name=your-pod namespace=you-ns process_name=your-process seconds=5
 
 | **Parameters:** see below. All parameters are optional except for ``pod_name`` and ``namespace``. ``pod_name`` can be the prefix of the pod name and doesn't need to be a full match.
 
@@ -127,7 +127,7 @@ http_stress_test
 
 .. code-block:: bash
 
-   robusta trigger http_stress_test url=http://grafana.default.svc:3000 slack_channel=test1 n=1000
+   robusta trigger http_stress_test url=http://grafana.default.svc:3000 n=1000
 
 | **Enabling it:** add to active_playbooks.yaml before manually trigger using the Robusta CLI (as described above):
 

docs/user-guide/playbook-configuration.rst

@@ -30,6 +30,63 @@ Here is a sample ``active_playbooks.yaml`` which enables two playbooks:
      - name: "python_profiler"
      - name: "restart_loop_reporter"
 
+Playbooks sinks
+^^^^^^^^^^^^^^^^^^
+| With Robusta, playbooks results can be send to one or more sinks.
+| Currently 3 sink types are supported:
+* *slack:* - Send playbooks results to Slack channel
+* *robusta:* - Send playbooks results Robusta's dedicated UI
+* *kafka:* - Send playbooks results Robusta's dedicated UI
+
+| To use sinks, first define the available named sinks.
+| For example: (in the ``active_playbooks.yaml`` file)
+
+.. code-block:: yaml
+
+    sinks_config:
+    - sink_name: "robusta ui"
+      sink_type: "robusta"
+      params:
+        account_id: "MY ROBUSTA ACCOUNT ID"
+        store_url: "MY ROBUSTA STORE URL"
+        api_key: "MY ROBUSTA ACCOUNT API KEY"
+    - sink_name: "alerts slack"
+      sink_type: "slack"
+      params:
+        slack_channel: "robusta alerts channel"
+    - sink_name: "my kafka sink"
+      sink_type: "kafka"
+      params:
+        kafka_url: "localhost:9092"
+        topic: "robusta-playbooks"
+
+
+| **Note:** Create your Robusta account, to get the parameters for the Robusta sink `here <https://robusta.dev>`_ .
+
+| By default, all playbooks will forward the results to the default sinks.
+| The default sinks are defined in the ``global_config`` section.
+| For example:
+
+.. code-block:: yaml
+
+   global_config:
+    sinks:
+    - "robusta ui"
+    - "alerts slack"
+
+| The default sinks list can be overridden, per playbook:
+
+.. code-block:: yaml
+
+   active_playbooks
+     - name: "add_deployment_lines_to_grafana"
+       sinks:
+       - "my kafka sink"
+       action_params:
+         grafana_dashboard_uid: "uid_from_url"
+         grafana_api_key: "grafana_api_key_with_editor_role"
+         grafana_service_name: "grafana.namespace.svc.cluster.local:3000"
+
 
 Playbook parameters
 ^^^^^^^^^^^^^^^^^^^
@@ -98,6 +155,7 @@ To avoid repeating yourself you can define trigger_params and parameters globall
 .. code-block:: yaml
 
    global_config:
+     cluster_name: "my-staging-cluster"
      grafana_api_key: "grafana_api_key_with_editor_role"
      grafana_service_name: "grafana.namespace.svc.cluster.local:3000"
 
@@ -114,3 +172,4 @@ To avoid repeating yourself you can define trigger_params and parameters globall
        trigger_params:
          name_prefix: "App2"
 
+| **Note:** The ``cluster_name`` is a required parameter, since it's used for sinks as the cluster identifier. (``cluster_name`` should be unique among different clusters)

playbooks/active_playbooks.yaml

@@ -1,6 +1,19 @@
+sinks_config:
+- sink_name: "robusta platform"
+  sink_type: "robusta"
+  params:
+    account_id: "<ROBUSTA_ACCOUNT_ID>"
+    store_url: "https://xvyhpoxfmtpuqqeyqkso.supabase.co"
+    api_key: "<ROBUSTA_API_KEY>"
+- sink_name: "alerts slack"
+  sink_type: "slack"
+  params:
+    slack_channel: "<DEFAULT_SLACK_CHANNEL>"
 global_config:
-  slack_channel: "<DEFAULT_SLACK_CHANNEL>"
-
+  cluster_name: "<CLUSTER_NAME>"
+  sinks: # These are the default sinks. The named sinks list can be overridden, per playbook
+  - "alerts slack"
+#<ENABLE_ROBUSTA_SINK>
 active_playbooks:
   - name: "restart_loop_reporter"
     action_params: