mention another performance tuning trick

mojolicious · Jan 18, 2014 · acc842e · acc842e
1 parent c71e687
commit acc842e
Show file tree

Hide file tree

Showing 2 changed files with 7 additions and 5 deletions.
diff --git a/lib/Mojo/Server/Hypnotoad.pm b/lib/Mojo/Server/Hypnotoad.pm
@@ -279,9 +279,9 @@ Listen backlog size, defaults to C<SOMAXCONN>.
   clients => 100
 
 Maximum number of parallel client connections per worker process, defaults to
-C<1000>. Note that depending on how much your application may block, you might
-want to decrease this value and increase L</"workers"> instead for better
-performance.
+C<1000>. Note that depending on how many blocking operations your application
+may perform, you might want to decrease this value and increase L</"workers">
+instead for better performance.
 
 =head2 graceful_timeout
 
@@ -386,7 +386,8 @@ Username for worker processes.
 
 Number of worker processes, defaults to C<4>. A good rule of thumb is two
 worker processes per CPU core for applications that perform mostly
-non-blocking operations, blocking operations often require more.
+non-blocking operations, blocking operations often require more and profit
+from decreasing the number of concurrent L</"clients">.
 
 =head1 METHODS
 

diff --git a/lib/Mojo/Server/Prefork.pm b/lib/Mojo/Server/Prefork.pm
@@ -463,7 +463,8 @@ Full path of process id file, defaults to a random temporary path.
 
 Number of worker processes, defaults to C<4>. A good rule of thumb is two
 worker processes per CPU core for applications that perform mostly
-non-blocking operations, blocking operations often require more.
+non-blocking operations, blocking operations often require more and profit
+from decreasing the number of concurrent L<Mojo::Server::Daemon/"clients">.
 
 =head1 METHODS