[MEDIUM] implement bind-process to limit service presence by process
The "bind-process" keyword lets the admin select which instances may
run on which process (in multi-process mode). It makes it easier to
more evenly distribute the load across multiple processes by avoiding
having too many listen to the same IP:ports.
diff --git a/doc/configuration.txt b/doc/configuration.txt
index fd62ef1..b410a23 100644
--- a/doc/configuration.txt
+++ b/doc/configuration.txt
@@ -523,6 +523,7 @@
backlog X X X -
balance X - X X
bind - X X -
+bind-process X X X X
block - X X X
capture cookie - X X -
capture request header - X X -
@@ -903,6 +904,57 @@
See also : "source".
+bind-process [ all | odd | even | <number 1-32> ] ...
+ Limit visibility of an instance to a certain set of processes numbers.
+ May be used in sections : defaults | frontend | listen | backend
+ yes | yes | yes | yes
+ Arguments :
+ all All process will see this instance. This is the default. It
+ may be used to override a default value.
+
+ odd This instance will be enabled on processes 1,3,5,...31. This
+ option may be combined with other numbers.
+
+ even This instance will be enabled on processes 2,4,6,...32. This
+ option may be combined with other numbers. Do not use it
+ with less than 2 processes otherwise some instances might be
+ missing from all processes.
+
+ number The instance will be enabled on this process number, between
+ 1 and 32. You must be careful not to reference a process
+ number greater than the configured global.nbproc, otherwise
+ some instances might be missing from all processes.
+
+ This keyword limits binding of certain instances to certain processes. This
+ is useful in order not to have too many processes listening to the same
+ ports. For instance, on a dual-core machine, it might make sense to set
+ 'nbproc 2' in the global section, then distributes the listeners among 'odd'
+ and 'even' instances.
+
+ At the moment, it is not possible to reference more than 32 processes using
+ this keyword, but this should be more than enough for most setups. Please
+ note that 'all' really means all processes and is not limited to the first
+ 32.
+
+ If some backends are referenced by frontends bound to other processes, the
+ backend automatically inherits the frontend's processes.
+
+ Example :
+ listen app_ip1
+ bind 10.0.0.1:80
+ bind_process odd
+
+ listen app_ip2
+ bind 10.0.0.2:80
+ bind_process even
+
+ listen management
+ bind 10.0.0.3:80
+ bind_process 1 2 3 4
+
+ See also : "nbproc" in global section.
+
+
block { if | unless } <condition>
Block a layer 7 request if/unless a condition is matched
May be used in sections : defaults | frontend | listen | backend
diff --git a/include/types/proxy.h b/include/types/proxy.h
index cc50690..2f67791 100644
--- a/include/types/proxy.h
+++ b/include/types/proxy.h
@@ -258,6 +258,7 @@
int uuid; /* universally unique proxy ID, used for SNMP */
int next_svid; /* next server-id, used for SNMP */
unsigned int backlog; /* force the frontend's listen backlog */
+ unsigned int bind_proc; /* bitmask of processes using this proxy. 0 = all. */
};
struct switching_rule {
diff --git a/src/cfgparse.c b/src/cfgparse.c
index 5136e94..e427a6f 100644
--- a/src/cfgparse.c
+++ b/src/cfgparse.c
@@ -668,6 +668,7 @@
curproxy->state = defproxy.state;
curproxy->options = defproxy.options;
curproxy->options2 = defproxy.options2;
+ curproxy->bind_proc = defproxy.bind_proc;
curproxy->lbprm.algo = defproxy.lbprm.algo;
curproxy->except_net = defproxy.except_net;
curproxy->except_mask = defproxy.except_mask;
@@ -930,6 +931,39 @@
else if (!strcmp(args[0], "enabled")) { /* enables this proxy (used to revert a disabled default) */
curproxy->state = PR_STNEW;
}
+ else if (!strcmp(args[0], "bind-process")) { /* enable this proxy only on some processes */
+ int cur_arg = 1;
+ unsigned int set = 0;
+
+ while (*args[cur_arg]) {
+ int u;
+ if (strcmp(args[cur_arg], "all") == 0) {
+ set = 0;
+ break;
+ }
+ else if (strcmp(args[cur_arg], "odd") == 0) {
+ set |= 0x55555555;
+ }
+ else if (strcmp(args[cur_arg], "even") == 0) {
+ set |= 0xAAAAAAAA;
+ }
+ else {
+ u = str2uic(args[cur_arg]);
+ if (u < 1 || u > 32) {
+ Alert("parsing [%s:%d]: %s expects 'all', 'odd', 'even', or process numbers from 1 to 32.\n",
+ file, linenum, args[0]);
+ return -1;
+ }
+ if (u > global.nbproc) {
+ Warning("parsing [%s:%d]: %s references process number higher than global.nbproc.\n",
+ file, linenum, args[0]);
+ }
+ set |= 1 << (u - 1);
+ }
+ cur_arg++;
+ }
+ curproxy->bind_proc = set;
+ }
else if (!strcmp(args[0], "acl")) { /* add an ACL */
if (curproxy == &defproxy) {
Alert("parsing [%s:%d] : '%s' not allowed in 'defaults' section.\n", file, linenum, args[0]);
@@ -3168,6 +3202,11 @@
} else {
free(curproxy->defbe.name);
curproxy->defbe.be = target;
+ /* we force the backend to be present on at least all of
+ * the frontend's processes.
+ */
+ target->bind_proc = curproxy->bind_proc ?
+ (target->bind_proc | curproxy->bind_proc) : 0;
}
}
@@ -3193,6 +3232,11 @@
} else {
free((void *)exp->replace);
exp->replace = (const char *)target;
+ /* we force the backend to be present on at least all of
+ * the frontend's processes.
+ */
+ target->bind_proc = curproxy->bind_proc ?
+ (target->bind_proc | curproxy->bind_proc) : 0;
}
}
}
@@ -3214,6 +3258,11 @@
} else {
free((void *)rule->be.name);
rule->be.backend = target;
+ /* we force the backend to be present on at least all of
+ * the frontend's processes.
+ */
+ target->bind_proc = curproxy->bind_proc ?
+ (target->bind_proc | curproxy->bind_proc) : 0;
}
}
diff --git a/src/haproxy.c b/src/haproxy.c
index dbb074d..cc6864e 100644
--- a/src/haproxy.c
+++ b/src/haproxy.c
@@ -1106,6 +1106,7 @@
}
if (global.mode & MODE_DAEMON) {
+ struct proxy *px;
int ret = 0;
int proc;
@@ -1131,6 +1132,16 @@
free(global.pidfile);
global.pidfile = NULL;
+ /* we might have to unbind some proxies from some processes */
+ px = proxy;
+ while (px != NULL) {
+ if (px->bind_proc && px->state != PR_STSTOPPED) {
+ if (!(px->bind_proc & (1 << proc)))
+ stop_proxy(px);
+ }
+ px = px->next;
+ }
+
if (proc == global.nbproc)
exit(0); /* parent must leave */