Run two nova-api and -metadata processes per pods

We want our services to be scaled by pod replicas but we still need two processes per API pods to always have a room for a healthcheck API query.
openstack-k8s-operators · Sep 7, 2023 · 8cf10ad · 8cf10ad
1 parent 480fa3a
commit 8cf10ad
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 2 deletions.
diff --git a/templates/novaapi/config/httpd.conf b/templates/novaapi/config/httpd.conf
@@ -38,7 +38,9 @@ LogLevel info
   WSGIProcessGroup nova-api
   WSGIApplicationGroup %{GLOBAL}
   WSGIPassAuthorization On
-  WSGIDaemonProcess nova-api processes=1 threads=1 user=nova group=nova display-name=nova-api
+  ## In general we want nova-api to scale via k8s replicas but we need
+  ## two processes per replica to always has a room for a healthecheck query
+  WSGIDaemonProcess nova-api processes=2 threads=1 user=nova group=nova display-name=nova-api
   WSGIScriptAlias / /usr/bin/nova-api-wsgi
 </VirtualHost>
 

diff --git a/templates/novametadata/config/httpd.conf b/templates/novametadata/config/httpd.conf
@@ -38,7 +38,9 @@ LogLevel info
   WSGIProcessGroup nova-metadata
   WSGIApplicationGroup %{GLOBAL}
   WSGIPassAuthorization On
-  WSGIDaemonProcess nova-metadata processes=1 threads=1 user=nova group=nova display-name=nova-metadata-api
+  ## In general we want nova-metadata to scale via k8s replicas but we need
+  ## two processes per replica to always has a room for a healthecheck query
+  WSGIDaemonProcess nova-metadata processes=2 threads=1 user=nova group=nova display-name=nova-metadata-api
   WSGIScriptAlias / /usr/bin/nova-metadata-wsgi
 </VirtualHost>