2 # IP Virtual Server configuration
5 tristate "IP virtual server support (EXPERIMENTAL)"
8 IP Virtual Server support will let you build a high-performance
9 virtual server based on cluster of two or more real servers. This
10 option must be enabled for at least one of the clustered computers
11 that will take care of intercepting incoming connections to a
12 single IP address and scheduling them to real servers.
14 Three request dispatching techniques are implemented, they are
15 virtual server via NAT, virtual server via tunneling and virtual
16 server via direct routing. The several scheduling algorithms can
17 be used to choose which server the connection is directed to,
18 thus load balancing can be achieved among the servers. For more
19 information and its administration program, please visit the
20 following URL: <http://www.linuxvirtualserver.org/>.
22 If you want to compile it in kernel, say Y. To compile it as a
23 module, choose M here. If unsure, say N.
28 bool "IPv6 support for IPVS (DANGEROUS)"
29 depends on EXPERIMENTAL && (IPV6 = y || IP_VS = IPV6)
31 Add IPv6 support to IPVS. This is incomplete and might be dangerous.
36 bool "IP virtual server debugging"
38 Say Y here if you want to get additional messages useful in
39 debugging the IP virtual server code. You can change the debug
40 level in /proc/sys/net/ipv4/vs/debug_level
43 int "IPVS connection table size (the Nth power of 2)"
46 The IPVS connection hash table uses the chaining scheme to handle
47 hash collisions. Using a big IPVS connection hash table will greatly
48 reduce conflicts when there are hundreds of thousands of connections
51 Note the table size must be power of 2. The table size will be the
52 value of 2 to the your input number power. The number to choose is
53 from 8 to 20, the default number is 12, which means the table size
54 is 4096. Don't input the number too small, otherwise you will lose
55 performance on it. You can adapt the table size yourself, according
56 to your virtual server application. It is good to set the table size
57 not far less than the number of connections per second multiplying
58 average lasting time of connection in the table. For example, your
59 virtual server gets 200 connections per second, the connection lasts
60 for 200 seconds in average in the connection table, the table size
61 should be not far less than 200x200, it is good to set the table
64 Another note that each connection occupies 128 bytes effectively and
65 each hash entry uses 8 bytes, so you can estimate how much memory is
68 comment "IPVS transport protocol load balancing support"
70 config IP_VS_PROTO_TCP
71 bool "TCP load balancing support"
73 This option enables support for load balancing TCP transport
74 protocol. Say Y if unsure.
76 config IP_VS_PROTO_UDP
77 bool "UDP load balancing support"
79 This option enables support for load balancing UDP transport
80 protocol. Say Y if unsure.
82 config IP_VS_PROTO_AH_ESP
86 config IP_VS_PROTO_ESP
87 bool "ESP load balancing support"
88 select IP_VS_PROTO_AH_ESP
90 This option enables support for load balancing ESP (Encapsulation
91 Security Payload) transport protocol. Say Y if unsure.
94 bool "AH load balancing support"
95 select IP_VS_PROTO_AH_ESP
97 This option enables support for load balancing AH (Authentication
98 Header) transport protocol. Say Y if unsure.
100 comment "IPVS scheduler"
103 tristate "round-robin scheduling"
105 The robin-robin scheduling algorithm simply directs network
106 connections to different real servers in a round-robin manner.
108 If you want to compile it in kernel, say Y. To compile it as a
109 module, choose M here. If unsure, say N.
112 tristate "weighted round-robin scheduling"
114 The weighted robin-robin scheduling algorithm directs network
115 connections to different real servers based on server weights
116 in a round-robin manner. Servers with higher weights receive
117 new connections first than those with less weights, and servers
118 with higher weights get more connections than those with less
119 weights and servers with equal weights get equal connections.
121 If you want to compile it in kernel, say Y. To compile it as a
122 module, choose M here. If unsure, say N.
125 tristate "least-connection scheduling"
127 The least-connection scheduling algorithm directs network
128 connections to the server with the least number of active
131 If you want to compile it in kernel, say Y. To compile it as a
132 module, choose M here. If unsure, say N.
135 tristate "weighted least-connection scheduling"
137 The weighted least-connection scheduling algorithm directs network
138 connections to the server with the least active connections
139 normalized by the server weight.
141 If you want to compile it in kernel, say Y. To compile it as a
142 module, choose M here. If unsure, say N.
145 tristate "locality-based least-connection scheduling"
147 The locality-based least-connection scheduling algorithm is for
148 destination IP load balancing. It is usually used in cache cluster.
149 This algorithm usually directs packet destined for an IP address to
150 its server if the server is alive and under load. If the server is
151 overloaded (its active connection numbers is larger than its weight)
152 and there is a server in its half load, then allocate the weighted
153 least-connection server to this IP address.
155 If you want to compile it in kernel, say Y. To compile it as a
156 module, choose M here. If unsure, say N.
159 tristate "locality-based least-connection with replication scheduling"
161 The locality-based least-connection with replication scheduling
162 algorithm is also for destination IP load balancing. It is
163 usually used in cache cluster. It differs from the LBLC scheduling
164 as follows: the load balancer maintains mappings from a target
165 to a set of server nodes that can serve the target. Requests for
166 a target are assigned to the least-connection node in the target's
167 server set. If all the node in the server set are over loaded,
168 it picks up a least-connection node in the cluster and adds it
169 in the sever set for the target. If the server set has not been
170 modified for the specified time, the most loaded node is removed
171 from the server set, in order to avoid high degree of replication.
173 If you want to compile it in kernel, say Y. To compile it as a
174 module, choose M here. If unsure, say N.
177 tristate "destination hashing scheduling"
179 The destination hashing scheduling algorithm assigns network
180 connections to the servers through looking up a statically assigned
181 hash table by their destination IP addresses.
183 If you want to compile it in kernel, say Y. To compile it as a
184 module, choose M here. If unsure, say N.
187 tristate "source hashing scheduling"
189 The source hashing scheduling algorithm assigns network
190 connections to the servers through looking up a statically assigned
191 hash table by their source IP addresses.
193 If you want to compile it in kernel, say Y. To compile it as a
194 module, choose M here. If unsure, say N.
197 tristate "shortest expected delay scheduling"
199 The shortest expected delay scheduling algorithm assigns network
200 connections to the server with the shortest expected delay. The
201 expected delay that the job will experience is (Ci + 1) / Ui if
202 sent to the ith server, in which Ci is the number of connections
203 on the ith server and Ui is the fixed service rate (weight)
206 If you want to compile it in kernel, say Y. To compile it as a
207 module, choose M here. If unsure, say N.
210 tristate "never queue scheduling"
212 The never queue scheduling algorithm adopts a two-speed model.
213 When there is an idle server available, the job will be sent to
214 the idle server, instead of waiting for a fast one. When there
215 is no idle server available, the job will be sent to the server
216 that minimize its expected delay (The Shortest Expected Delay
217 scheduling algorithm).
219 If you want to compile it in kernel, say Y. To compile it as a
220 module, choose M here. If unsure, say N.
222 comment 'IPVS application helper'
225 tristate "FTP protocol helper"
226 depends on IP_VS_PROTO_TCP
228 FTP is a protocol that transfers IP address and/or port number in
229 the payload. In the virtual server via Network Address Translation,
230 the IP address and port number of real servers cannot be sent to
231 clients in ftp connections directly, so FTP protocol helper is
232 required for tracking the connection and mangling it back to that of
235 If you want to compile it in kernel, say Y. To compile it as a
236 module, choose M here. If unsure, say N.