From c84c8057a096ee0eb775a0dd1565ad3d34e5c474 Mon Sep 17 00:00:00 2001
From: Sebastien Robin <seb@nexedi.com>
Date: Thu, 8 Sep 2011 18:41:31 +0200
Subject: [PATCH] optimize haproxy server configuration and explain values
---
slapos/recipe/erp5/__init__.py | 12 +++++++++++-
1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/slapos/recipe/erp5/__init__.py b/slapos/recipe/erp5/__init__.py
index f9c478a1b..833b49f62 100644
--- a/slapos/recipe/erp5/__init__.py
+++ b/slapos/recipe/erp5/__init__.py
@@ -546,8 +546,18 @@ class Recipe(BaseSlapRecipe):
}
def installHaproxy(self, ip, port, name, server_check_path, url_list):
- server_template = """ server %(name)s %(address)s cookie %(name)s check inter 20s rise 2 fall 4"""
+ # inter must be quite short in order to detect quickly an unresponsive node
+ # and to detect quickly a node which is back
+ # rise must be minimal possible : 1, indeed, a node which is back don't need
+ # to sleep more time and we can give him work immediately
+ # fall should be quite sort. with inter at 3, and fall at 2, a node will be
+ # considered as dead after 6 seconds.
+ # maxconn should be set as the maximum thread we have per zope, like this
+ # haproxy will manage the queue of request with the possibility to
+ # move a request to another node if the initially selected one is dead
+ server_template = """ server %(name)s %(address)s cookie %(name)s check inter 3s rise 1 fall 2 maxconn %(cluster_zope_thread_amount)s""" %
config = dict(name=name, ip=ip, port=port,
+ cluster_zope_thread_amount=self.options.get('cluster_zope_thread_amount', 1),
server_check_path=server_check_path,)
i = 1
server_list = []
--
2.30.9