leandrof/linux
1
0
Fork 0
forked from Minki/linux
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ipvs: add sysctl_run_estimation to support disable estimation

Browse Source 
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: Dust Li <dust.li@linux.alibaba.com>
Acked-by: Julian Anastasov <ja@ssi.bg>
Acked-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This commit is contained in:
Dust Li 2021-08-20 13:37:52 +08:00 committed by Pablo Neira Ayuso
parent c514fbb623
commit 2232642ec3
4 changed files with 35 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
Documentation/networking/ipvs-sysctl.rst
View File

@ -300,3 +300,14 @@ sync_version - INTEGER
Kernels with this sync_version entry are able to receive messages Kernels with this sync_version entry are able to receive messages
of both version 1 and version 2 of the synchronisation protocol. 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.

11
include/net/ip_vs.h
View File

@ -931,6 +931,7 @@ struct netns_ipvs {
int sysctl_conn_reuse_mode; int sysctl_conn_reuse_mode;
int sysctl_schedule_icmp; int sysctl_schedule_icmp;
int sysctl_ignore_tunneled; int sysctl_ignore_tunneled;
int sysctl_run_estimation;
/* ip_vs_lblc */ /* ip_vs_lblc */
int sysctl_lblc_expiration; int sysctl_lblc_expiration;
@ -1071,6 +1072,11 @@ static inline int sysctl_cache_bypass(struct netns_ipvs *ipvs)
return ipvs->sysctl_cache_bypass; return ipvs->sysctl_cache_bypass;
} }
static inline int sysctl_run_estimation(struct netns_ipvs *ipvs)
{
return ipvs->sysctl_run_estimation;
}
#else #else
static inline int sysctl_sync_threshold(struct netns_ipvs *ipvs) 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; return 0;
} }
static inline int sysctl_run_estimation(struct netns_ipvs *ipvs)
{
return 1;
}
#endif #endif
/* IPVS core functions /* IPVS core functions

8
net/netfilter/ipvs/ip_vs_ctl.c
View File

@ -2017,6 +2017,12 @@ static struct ctl_table vs_vars[] = {
.mode = 0644, .mode = 0644,
.proc_handler = proc_dointvec, .proc_handler = proc_dointvec,
}, },
{
.procname = "run_estimation",
.maxlen = sizeof(int),
.mode = 0644,
.proc_handler = proc_dointvec,
},
#ifdef CONFIG_IP_VS_DEBUG #ifdef CONFIG_IP_VS_DEBUG
{ {
.procname = "debug_level", .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_conn_reuse_mode;
tbl[idx++].data = &ipvs->sysctl_schedule_icmp; tbl[idx++].data = &ipvs->sysctl_schedule_icmp;
tbl[idx++].data = &ipvs->sysctl_ignore_tunneled; 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); ipvs->sysctl_hdr = register_net_sysctl(net, "net/ipv4/vs", tbl);
if (ipvs->sysctl_hdr == NULL) { if (ipvs->sysctl_hdr == NULL) {

5
net/netfilter/ipvs/ip_vs_est.c
View File

@ -100,6 +100,9 @@ static void estimation_timer(struct timer_list *t)
u64 rate; u64 rate;
struct netns_ipvs *ipvs = from_timer(ipvs, t, est_timer); struct netns_ipvs *ipvs = from_timer(ipvs, t, est_timer);
if (!sysctl_run_estimation(ipvs))
goto skip;
spin_lock(&ipvs->est_lock); spin_lock(&ipvs->est_lock);
list_for_each_entry(e, &ipvs->est_list, list) { list_for_each_entry(e, &ipvs->est_list, list) {
s = container_of(e, struct ip_vs_stats, est); 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(&s->lock);
} }
spin_unlock(&ipvs->est_lock); spin_unlock(&ipvs->est_lock);
skip:
mod_timer(&ipvs->est_timer, jiffies + 2*HZ); mod_timer(&ipvs->est_timer, jiffies + 2*HZ);
} }