Performance and Tuning - Revision history

Dogonthesun: /* Connection Hash Table Size */

2012-01-26T11:37:51Z

‎Connection Hash Table Size

Dogonthesun: /* Connection Hash Table Size */

2012-01-26T11:36:45Z

‎Connection Hash Table Size

Wensong: /* Tools */

2008-11-07T06:53:56Z

‎Tools

Wensong: /* Tools */

2008-10-23T12:51:54Z

‎Tools

Wensong: /* Tools */

2008-10-23T12:31:31Z

‎Tools

Wensong at 14:54, 10 November 2006

2006-11-10T14:54:20Z

Wensong at 14:29, 4 September 2006

2006-09-04T14:29:11Z

Wensong: /* Netfilter Connection Track */

2006-05-13T15:41:15Z

‎Netfilter Connection Track

Wensong: /* Connection Hash Table Size */

2006-05-13T15:28:58Z

‎Connection Hash Table Size

Wensong: /* Performance Tuning */

2006-05-13T14:41:42Z

‎Performance Tuning

@@ Line 38: / Line 38: @@
    Network packet filtering framework (Netfilter)  --->
      IP: Virtual Server Configuration -->
-       <M> virtual server support (EXPERIMENTAL)
+       [*]   IPv6 support for IPVS
        [ ]   IP virtual server debugging
-       (12)   IPVS connection table size (the Nth power of 2)
+       (20)  IPVS connection table size (the Nth power of 2)
-       --- IPVS scheduler
+            *** IPVS transport protocol load balancing support ***
-       <M>   round-robin scheduling
+      [*]   TCP load balancing support
-       <M>   weighted round-robin scheduling
+      [*]   UDP load balancing support
-       <M>   least-connection scheduling scheduling
+      [*]   ESP load balancing support
-       <M>   weighted least-connection scheduling
+      [*]   AH load balancing support
-       <M>   locality-based least-connection scheduling
+            *** IPVS scheduler ***
-       <M>   locality-based least-connection with replication scheduling
+       <M>   round-robin scheduling
-       <M>   destination hashing scheduling
+       <M>   weighted round-robin scheduling
-       <M>   source hashing scheduling
+       <M>   least-connection scheduling
-       <M>   shortest expected delay scheduling
+       <M>   weighted least-connection scheduling
-       <M>   never queue scheduling
+       <M>   locality-based least-connection scheduling
-      --- IPVS application helper
+       <M>   locality-based least-connection with replication scheduling
-       <M>   FTP protocol helper
+       <M>   destination hashing scheduling
 </pre>

@@ Line 36: / Line 36: @@
 <pre>
 Networking Options -->
-   IP: Virtual Server Configuration -->
+   Network packet filtering framework (Netfilter)  --->
-    <M> virtual server support (EXPERIMENTAL)
+    IP: Virtual Server Configuration -->
-    [ ]   IP virtual server debugging
+      <M> virtual server support (EXPERIMENTAL)
-    (12)   IPVS connection table size (the Nth power of 2)
+      [ ]   IP virtual server debugging
-    --- IPVS scheduler
+      (12)   IPVS connection table size (the Nth power of 2)
-    <M>   round-robin scheduling
+      --- IPVS scheduler
-    <M>   weighted round-robin scheduling
+      <M>   round-robin scheduling
-    <M>   least-connection scheduling scheduling
+      <M>   weighted round-robin scheduling
-    <M>   weighted least-connection scheduling
+      <M>   least-connection scheduling scheduling
-    <M>   locality-based least-connection scheduling
+      <M>   weighted least-connection scheduling
-    <M>   locality-based least-connection with replication scheduling
+      <M>   locality-based least-connection scheduling
-    <M>   destination hashing scheduling
+      <M>   locality-based least-connection with replication scheduling
-    <M>   source hashing scheduling
+      <M>   destination hashing scheduling
-    <M>   shortest expected delay scheduling
+      <M>   source hashing scheduling
-    <M>   never queue scheduling
+      <M>   shortest expected delay scheduling
-    --- IPVS application helper
+      <M>   never queue scheduling
-    <M>   FTP protocol helper
+      --- IPVS application helper
 </pre>

@@ Line 6: / Line 6: @@
 * ab: apache benchmark
 * httperf: a tool for measuring web server performance from HP, see [http://www.hpl.hp.com/research/linux/httperf/ the httperf homepage]
 * specweb: web server benchmark by spec.org, it is now [http://www.spec.org/web2005/ specweb2005].
 * webbench: a licensed benchmark program that measures the performance of Web servers, developed by VeriTest.

@@ Line 3: / Line 3: @@
 === Tools ===
 * ab: apache benchmark
 * httperf: a tool for measuring web server performance from HP, see [http://www.hpl.hp.com/research/linux/httperf/ the httperf homepage]

@@ Line 4: / Line 4: @@
 * ab: apache benchmark
 * specweb: web server benchmark by spec.org, it is now [http://www.spec.org/web2005/ specweb2005].
 * webbench: a licensed benchmark program that measures the performance of Web servers, developed by VeriTest.
 == Performance Tuning ==

← Older revision		Revision as of 14:54, 10 November 2006
Line 71:		Line 71:

	* [http://www.kegel.com/c10k.html The C10K problem] written by Dan Kegel, good notes on how to configure operating systems and write code to support 10K clients on a single server		* [http://www.kegel.com/c10k.html The C10K problem] written by Dan Kegel, good notes on how to configure operating systems and write code to support 10K clients on a single server
		+
		+	[[Category:LVS Handbook]]

@@ Line 58: / Line 58: @@
 Julian compared the performance of IPVS with ip_conntrack and without ip_conntrack. See http://archive.linuxvirtualserver.org/html/lvs-users/2001-12/msg00141.html
 == External Links ==
 * [http://www.kegel.com/c10k.html The C10K problem] written by Dan Kegel, good notes on how to configure operating systems and write code to support 10K clients on a single server

← Older revision		Revision as of 15:28, 13 May 2006
Line 11:		Line 11:

	=== Connection Hash Table Size ===		=== Connection Hash Table Size ===
		+
		+	The IPVS connection hash table uses the chaining scheme to handle
		+	hash collisions. Using a big IPVS connection hash table will greatly
		+	reduce conflicts when there are hundreds of thousands of connections
		+	in the hash table.
		+
		+	Note the table size must be power of 2. The table size will be the
		+	value of 2 to the your input number power. The number to choose is
		+	from 8 to 20, the default number is 12, which means the table size
		+	is 4096. Don't input the number too small, otherwise you will lose
		+	performance on it. You can adapt the table size yourself, according
		+	to your virtual server application. It is good to set the table size
		+	not far less than the number of connections per second multiplying
		+	average lasting time of connection in the table. For example, your
		+	virtual server gets 200 connections per second, the connection lasts
		+	for 200 seconds in average in the connection table, the table size
		+	should be not far less than 200x200, it is good to set the table
		+	size 32768 (2**15).
		+
		+	We can configure the size of IPVS connection hash table before compiling the Linux kernel. Here are the IPVS configurations in the 'make menuconfig' menu:
		+
		+	<pre>
		+	Networking Options -->
		+	IP: Virtual Server Configuration -->
		+	<M> virtual server support (EXPERIMENTAL)
		+	[ ] IP virtual server debugging
		+	(12) IPVS connection table size (the Nth power of 2)
		+	--- IPVS scheduler
		+	<M> round-robin scheduling
		+	<M> weighted round-robin scheduling
		+	<M> least-connection scheduling scheduling
		+	<M> weighted least-connection scheduling
		+	<M> locality-based least-connection scheduling
		+	<M> locality-based least-connection with replication scheduling
		+	<M> destination hashing scheduling
		+	<M> source hashing scheduling
		+	<M> shortest expected delay scheduling
		+	<M> never queue scheduling
		+	--- IPVS application helper
		+	<M> FTP protocol helper
		+	</pre>

	=== Netfilter Connection Track ===		=== Netfilter Connection Track ===