Commit 2232642e authored by Dust Li's avatar Dust Li Committed by Pablo Neira Ayuso
Browse files

ipvs: add sysctl_run_estimation to support disable estimation 

estimation_timer will iterate the est_list to do estimation
for each ipvs stats. When there are lots of services, the
list can be very large.
We found that estimation_timer() run for more then 200ms on a
machine with 104 CPU and 50K services.

yunhong-cgl jiang report the same phenomenon before:
https://www.spinics.net/lists/lvs-devel/msg05426.html



In some cases(for example a large K8S cluster with many ipvs services),
ipvs estimation may not be needed. So adding a sysctl blob to allow
users to disable this completely.

Default is: 1 (enable)

Cc: yunhong-cgl jiang <xintian1976@gmail.com>
Signed-off-by: default avatarDust Li <dust.li@linux.alibaba.com>
Acked-by: default avatarJulian Anastasov <ja@ssi.bg>
Acked-by: default avatarSimon Horman <horms@verge.net.au>
Signed-off-by: default avatarPablo Neira Ayuso <pablo@netfilter.org>
parent c514fbb6

Documentation/networking/ipvs-sysctl.rst

+11 −0
Original line number Diff line number Diff line
@@ -300,3 +300,14 @@ sync_version - INTEGER 


	Kernels with this sync_version entry are able to receive messages

	of both version 1 and version 2 of the synchronisation protocol.



run_estimation - BOOLEAN

	0 - disabled

	not 0 - enabled (default)



	If disabled, the estimation will be stop, and you can't see

	any update on speed estimation data.



	You can always re-enable estimation by setting this value to 1.

	But be careful, the first estimation after re-enable is not

	accurate.

include/net/ip_vs.h

+11 −0
Original line number Diff line number Diff line
@@ -931,6 +931,7 @@ struct netns_ipvs { 
	int			sysctl_conn_reuse_mode;

	int			sysctl_schedule_icmp;

	int			sysctl_ignore_tunneled;

	int			sysctl_run_estimation;



	/* ip_vs_lblc */

	int			sysctl_lblc_expiration;
@@ -1071,6 +1072,11 @@ static inline int sysctl_cache_bypass(struct netns_ipvs *ipvs) 
	return ipvs->sysctl_cache_bypass;

}



static inline int sysctl_run_estimation(struct netns_ipvs *ipvs)

{

	return ipvs->sysctl_run_estimation;

}



#else



static inline int sysctl_sync_threshold(struct netns_ipvs *ipvs)
@@ -1163,6 +1169,11 @@ static inline int sysctl_cache_bypass(struct netns_ipvs *ipvs) 
	return 0;

}



static inline int sysctl_run_estimation(struct netns_ipvs *ipvs)

{

	return 1;

}



#endif



/* IPVS core functions

net/netfilter/ipvs/ip_vs_ctl.c

+8 −0
Original line number Diff line number Diff line
@@ -2017,6 +2017,12 @@ static struct ctl_table vs_vars[] = { 
		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "run_estimation",

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

#ifdef CONFIG_IP_VS_DEBUG

	{

		.procname	= "debug_level",
@@ -4090,6 +4096,8 @@ static int __net_init ip_vs_control_net_init_sysctl(struct netns_ipvs *ipvs) 
	tbl[idx++].data = &ipvs->sysctl_conn_reuse_mode;

	tbl[idx++].data = &ipvs->sysctl_schedule_icmp;

	tbl[idx++].data = &ipvs->sysctl_ignore_tunneled;

	ipvs->sysctl_run_estimation = 1;

	tbl[idx++].data = &ipvs->sysctl_run_estimation;



	ipvs->sysctl_hdr = register_net_sysctl(net, "net/ipv4/vs", tbl);

	if (ipvs->sysctl_hdr == NULL) {

net/netfilter/ipvs/ip_vs_est.c

+5 −0
Original line number Diff line number Diff line
@@ -100,6 +100,9 @@ static void estimation_timer(struct timer_list *t) 
	u64 rate;

	struct netns_ipvs *ipvs = from_timer(ipvs, t, est_timer);



	if (!sysctl_run_estimation(ipvs))

		goto skip;



	spin_lock(&ipvs->est_lock);

	list_for_each_entry(e, &ipvs->est_list, list) {

		s = container_of(e, struct ip_vs_stats, est);
@@ -131,6 +134,8 @@ static void estimation_timer(struct timer_list *t) 
		spin_unlock(&s->lock);

	}

	spin_unlock(&ipvs->est_lock);



skip:

	mod_timer(&ipvs->est_timer, jiffies + 2*HZ);

}