Make deployments configurable (#59)

* Update charts and configuration to optionally disable services * Fix default configuration Fix typos from #56 * Update charts to support multiple neon-core ingresses Update CLI entrypoint to make iris Gradio UI deployments optional * Add prompt to set image tag in core setup * Add method to read configured backend domain for default iris endpoint * Add LLMs to backend configuration checks * Update chart versioning and push to Helm repo * Update tests to account for added LLMs * Update logging to clearly print disabled MQ and HTTP backend services --------- Co-authored-by: Daniel McKnight <[email protected]>
NeonGeckoCom · Jan 23, 2024 · 77ea21d · 77ea21d
1 parent 9290ab0
commit 77ea21d
Show file tree

Hide file tree

Showing 24 changed files with 292 additions and 100 deletions.
diff --git a/neon_diana_utils/configuration.py b/neon_diana_utils/configuration.py
@@ -32,7 +32,7 @@
 
 from enum import Enum
 from pprint import pformat
-from typing import Optional
+from typing import Optional, Set
 from os import makedirs, listdir
 from os.path import expanduser, join, abspath, isfile, isdir, dirname
 from ovos_utils.xdg_utils import xdg_config_home
@@ -503,6 +503,61 @@ def _get_chatbots_mq_config(rmq_config: str) -> dict:
     return {"MQ": chatbot_config}
 
 
+def _get_unconfigured_mq_backend_services(config: dict) -> Set[str]:
+    """
+    Get a list of MQ Backend services that are not configured to run
+    @param config: dict Configuration (diana.yaml)
+    """
+    config_to_service = {'keys.api_services': 'neon-api-proxy',
+                         'keys.emails': 'neon-email-proxy',
+                         'keys.track_my_brands': 'neon-brands-service',
+                         'LLM_CHAT_GPT': 'neon-llm-chatgpt',
+                         'LLM_FASTCHAT': 'neon-llm-fastchat',
+                         'LLM_CLAUDE': 'neon-llm-claude',
+                         'LLM_GEMINI': 'neon-llm-gemini',
+                         'LLM_PALM2': 'neon-llm-palm'}
+    disabled = list()
+    for key, service in config_to_service.items():
+        if '.' in key:
+            parts = key.split('.')
+            test = dict(config)
+            for part in parts:
+                test = test.get(part)
+            if not test:
+                disabled.append(service)
+        else:
+            if not config.get(key):
+                disabled.append(service)
+    return set(disabled)
+
+
+def _get_optional_http_backend() -> Set[str]:
+    """
+    Get a set of optional HTTP backend services
+    """
+    return {'tts-larynx', 'tts-ljspeech', 'tts-mozilla', 'tts-nancy',
+            'tts-glados', 'ww-snowboy'}
+
+
+def _read_backend_domain(deploy_path: str) -> str:
+    """
+    Read backend configuration to determine the configured root domain
+    @param deploy_path: Path to diana-backend deployment
+    """
+    if isdir(join(deploy_path, "diana-backend")):
+        deploy_path = join(deploy_path, "diana-backend")
+    values_file = join(deploy_path, "values.yaml")
+    if not isfile(values_file):
+        raise FileNotFoundError(values_file)
+    with open(values_file, 'r') as f:
+        backend_config = yaml.safe_load(f)
+    root_domain = backend_config.get('backend', {}).get('diana-http',
+                                                        {}).get('domain')
+    if not root_domain:
+        raise ValueError(f"Expected config not found in: {values_file}")
+    return root_domain
+
+
 def configure_backend(username: str = None,
                       password: str = None,
                       output_path: str = None,
@@ -523,6 +578,11 @@ def configure_backend(username: str = None,
         click.echo(f"Path exists: {output_path}")
         return
 
+    # Collect user inputs for service configuration
+    keys_config = make_keys_config(False)
+    disabled_mq_services = list(
+        _get_unconfigured_mq_backend_services(keys_config))
+
     if orchestrator == Orchestrator.KUBERNETES:
         shutil.copytree(join(dirname(__file__), "templates", "backend"),
                         join(output_path, "diana-backend"))
@@ -541,6 +601,8 @@ def configure_backend(username: str = None,
         else:
             # Define a default value so secret can be generated
             encoded_token = get_github_encoded_auth("", "")
+            to_disable = ['neon-brands-service', 'neon-script-parser']
+            disabled_mq_services += to_disable
         confirmed = False
         email = ''
         domain = ''
@@ -557,13 +619,30 @@ def configure_backend(username: str = None,
                                 'tag': tag}))
             confirmed = click.confirm("Is this configuration correct?")
 
+        click.echo(f"The following MQ services are disabled: "
+                   f"{disabled_mq_services}")
+
+        if click.confirm(f"Disable optional HTTP Services?"):
+            disabled_http_services = _get_optional_http_backend()
+            click.echo(f"The following HTTP services are disabled: "
+                       f"{disabled_http_services}")
+        else:
+            disabled_http_services = set()
+
         # Generate values.yaml with configured params
         values_file = join(output_path, "diana-backend", "values.yaml")
         with open(values_file, 'r') as f:
             helm_values = yaml.safe_load(f)
         helm_values['backend']['letsencrypt']['email'] = email
         helm_values['backend']['diana-http']['domain'] = domain
         helm_values['backend']['ghTokenEncoded'] = encoded_token
+        for service in disabled_mq_services:
+            LOG.debug(f"Disable {service}")
+            helm_values['backend']['diana-mq'][service]['replicaCount'] = 0
+        for service in disabled_http_services:
+            LOG.debug(f"Disable {service}")
+            helm_values['backend']['diana-http'].setdefault(service, dict())
+            helm_values['backend']['diana-http'][service]['replicaCount'] = 0
         for service in helm_values['backend']['diana-mq']:
             helm_values['backend']['diana-mq'][service]['image']['tag'] = \
                 tag
@@ -591,7 +670,6 @@ def configure_backend(username: str = None,
         click.echo(f"Generated auth for services: {set(mq_auth_config.keys())}")
 
         # Generate `diana.yaml` output
-        keys_config = make_keys_config(False)
         if keys_config.get("LLM_CHAT_GPT"):
             llm_config = make_llm_bot_config()
         else:
@@ -692,23 +770,51 @@ def configure_neon_core(mq_user: str = None,
         return
 
     # Prompt for IRIS Web UI configuration
-    confirmed = False
-    iris_domain = "iris.diana.k8s"  # TODO: Read from backend config
-    while not confirmed:
-        iris_domain = click.prompt("Hostname for Iris Web UI", type=str,
-                                   default=iris_domain)
-        confirmed = click.confirm(f"Is {iris_domain} correct?")
+    iris_domain = None
+    if click.confirm("Configure IRIS Gradio Web UI?"):
+        confirmed = False
+        try:
+            domain = _read_backend_domain(output_path)
+        except FileNotFoundError:
+            domain = "diana.k8s"
+        while not confirmed:
+            iris_domain = click.prompt("Hostname for Iris Gradio Web UI",
+                                       type=str, default=f"iris.{domain}")
+            confirmed = click.confirm(f"Is {iris_domain} correct?")
 
     if orchestrator == Orchestrator.KUBERNETES:
         shutil.copytree(join(dirname(__file__), "templates", "neon"),
                         join(output_path, "neon-core"))
         neon_config_file = join(output_path, "neon-core", "neon.yaml")
-        # TODO: Configure image tag to use
+
+        # Determine image tag to use
+        tag = 'latest'
+        confirmed = False
+        while not confirmed:
+            tag = click.prompt("Image tags to use for Neon Core Services",
+                               type=str, default=tag)
+            confirmed = click.confirm(f"Is `{tag}` correct?")
         values = join(output_path, "neon-core", "values.yaml")
         with open(values, "r") as f:
             config = yaml.safe_load(f)
-        config['iris']['subdomain'], config['iris']['domain'] =\
-            iris_domain.split('.', 1)
+        for service in {'neon-messagebus', 'neon-speech', 'neon-skills',
+                        'neon-audio', 'neon-enclosure', 'neon-gui',
+                        'iris-gradio'}:
+            config['core'][service]['image']['tag'] = tag
+
+        if iris_domain:
+            iris_subdomain, iris_domain = iris_domain.split('.', 1)
+            config['core']['domain'] = iris_domain
+            config['core']['ingress']['rules'].append(
+                {'host': iris_subdomain, 'serviceName': 'neon-core-iris',
+                 'servicePort': 7860})
+        else:
+            click.echo("iris Gradio UI disabled")
+            config['core']['iris-gradio']['replicaCount'] = 0
+
+        if not config['core']['ingress']['rules']:
+            click.echo('No HTTP ingress configured')
+            config['core']['ingress']['enabled'] = False
         with open(values, 'w') as f:
             yaml.safe_dump(config, f)
     elif orchestrator == Orchestrator.COMPOSE:
@@ -817,7 +923,7 @@ def configure_klat_chat(external_url: str = None,
 
     forward_www = False
     if subdomain != "www":
-        forward_www = click.prompt(f"Route www.{domain} traffic to Klat?")
+        forward_www = click.confirm(f"Route www.{domain} traffic to Klat?")
 
     # Get Libretranslate HTTP API URL
     libretranslate_url = "https://libretranslate.2022.us"
@@ -914,10 +1020,10 @@ def configure_klat_chat(external_url: str = None,
              'servicePort': 8010}
         ]
         if forward_www:
-            helm_values['klat']['ingress']['rules'].append({
+            helm_values['klat']['ingress']['rules'].append(
                 {'host': 'www', 'serviceName': 'klat-chat-client',
                  'servicePort': 8001}
-            })
+            )
         with open(join(output_path, "klat-chat", "values.yaml"), 'w') as f:
             yaml.safe_dump(helm_values, f)
     else:

diff --git a/neon_diana_utils/helm_charts/backend/diana-backend/Chart.yaml b/neon_diana_utils/helm_charts/backend/diana-backend/Chart.yaml
@@ -15,7 +15,7 @@ type: application
 # This is the chart version. This version number should be incremented each time you make changes
 # to the chart and its templates, including the app version.
 # Versions are expected to follow Semantic Versioning (https://semver.org/)
-version: 0.1.22
+version: 0.1.23
 
 # This is the version number of the application being deployed. This version number should be
 # incremented each time you make changes to the application. Versions are not expected to
@@ -32,8 +32,8 @@ dependencies:
     version: 11.13.0
     repository: https://charts.bitnami.com/bitnami
   - name: diana-http
-    version: 0.0.11
+    version: 0.0.12
     repository: file://../http-services
   - name: diana-mq
-    version: 0.0.16
+    version: 0.0.17
     repository: file://../mq-services
diff --git a/neon_diana_utils/helm_charts/backend/http-services/Chart.yaml b/neon_diana_utils/helm_charts/backend/http-services/Chart.yaml
@@ -3,7 +3,7 @@ name: diana-http
 description: Deploy DIANA HTTP Services
 
 type: application
-version: 0.0.11
+version: 0.0.12
 appVersion: "1.0.1a19"
 dependencies:
   - name: libretranslate
@@ -14,10 +14,10 @@ dependencies:
     alias: tts-coqui
     version: 0.0.5
     repository: file://../../http/tts-coqui
-#  - name: tts-glados
-#    alias: tts-glados
-#    version: 0.0.1
-#    repository: https://neongeckocom.github.io/neon-diana-utils
+  - name: tts-glados
+    alias: tts-glados
+    version: 0.0.1
+    repository: https://neongeckocom.github.io/neon-diana-utils
   - name: tts-larynx
     alias: tts-larynx
     version: 0.0.5

diff --git a/neon_diana_utils/helm_charts/backend/http-services/values.yaml b/neon_diana_utils/helm_charts/backend/http-services/values.yaml
@@ -17,12 +17,13 @@ tts-coqui:
   ingress:
     enabled: False
 #  image.tag:
-#tts-glados:
-#  subdomain: &host_glados glados
-#  servicePort: *default_port
-#  ingress:
-#    enabled: False
-#  image.tag:
+tts-glados:
+  replicaCount: 0
+  subdomain: &host_glados glados
+  servicePort: *default_port
+  ingress:
+    enabled: False
+  image.tag:
 tts-larynx:
   subdomain: &host_larynx larynx
   servicePort: *default_port
@@ -69,9 +70,9 @@ ingress:
     - host: *host_coqui
       serviceName: backend-coqui
       servicePort: *default_port
-#    - host: *host_glados
-#      serviceName: tts-glados
-#      servicePort: *default_port
+    - host: *host_glados
+      serviceName: tts-glados
+      servicePort: *default_port
     - host: *host_larynx
       serviceName: backend-larynx
       servicePort: *default_port

diff --git a/neon_diana_utils/helm_charts/backend/mq-services/Chart.yaml b/neon_diana_utils/helm_charts/backend/mq-services/Chart.yaml
@@ -3,7 +3,7 @@ name: diana-mq
 description: Deploy DIANA MQ Services
 
 type: application
-version: 0.0.16
+version: 0.0.17
 appVersion: "1.0.1a19"
 dependencies:
   - name: neon-api-proxy

diff --git a/neon_diana_utils/helm_charts/backend/mq-services/values.yaml b/neon_diana_utils/helm_charts/backend/mq-services/values.yaml
@@ -15,3 +15,9 @@ neon-metrics-service:
 neon-script-parser:
   image:
     tag: *tag
+neon-llm-chatgpt:
+  image:
+    tag: *tag
+neon-llm-fastchat:
+  image:
+    tag: *tag
diff --git a/neon_diana_utils/helm_charts/base/base-neon/Chart.yaml b/neon_diana_utils/helm_charts/base/base-neon/Chart.yaml
@@ -3,5 +3,5 @@ name: base-neon
 description: Library chart for basic Neon Core Services
 
 type: library
-version: 0.0.4
+version: 0.0.5
 appVersion: "1.0.1a19"
diff --git a/neon_diana_utils/helm_charts/base/base-neon/templates/_ingress.tpl b/neon_diana_utils/helm_charts/base/base-neon/templates/_ingress.tpl
@@ -0,0 +1,40 @@
+{{- define "base-neon.ingress" -}}
+{{ if .Values.ingress.enabled }}
+{{- $domain := .Values.domain -}}
+{{- if semverCompare ">=1.19-0" .Capabilities.KubeVersion.GitVersion -}}
+apiVersion: networking.k8s.io/v1
+{{- else if semverCompare ">=1.14-0" .Capabilities.KubeVersion.GitVersion -}}
+apiVersion: networking.k8s.io/v1beta1
+{{- else -}}
+apiVersion: extensions/v1beta1
+{{- end }}
+kind: Ingress
+metadata:
+  name: ingress-iris-http
+  annotations:
+    kubernetes.io/ingress.class: {{ .Values.ingress.ingressClassName }}
+    cert-manager.io/issuer: {{ .Values.ingress.certIssuer }}
+spec:
+  tls:
+    - secretName: {{ .Values.ingress.tlsSecretName }}
+      hosts:
+        {{- range .Values.ingress.rules }}
+        - {{ printf "%s.%s " .host $.Values.domain }}
+        {{- end }}
+
+  rules:
+    {{- range .Values.ingress.rules }}
+    - host: {{ printf "%s.%s " .host $.Values.domain }}
+      http:
+        paths:
+          - path: /
+            pathType: Prefix
+            backend:
+              service:
+                name: {{ .serviceName }}
+                port:
+                  number: {{ .servicePort }}
+    {{- end }}
+
+  {{- end -}}
+{{- end -}}