diff options
-rw-r--r-- | Documentation/networking/ipvs-sysctl.txt | 13 | ||||
-rw-r--r-- | include/net/ip_vs.h | 11 | ||||
-rw-r--r-- | net/netfilter/ipvs/ip_vs_ctl.c | 7 | ||||
-rw-r--r-- | net/netfilter/ipvs/ip_vs_sync.c | 12 |
4 files changed, 43 insertions, 0 deletions
diff --git a/Documentation/networking/ipvs-sysctl.txt b/Documentation/networking/ipvs-sysctl.txt index 9573d0c..7a3c047 100644 --- a/Documentation/networking/ipvs-sysctl.txt +++ b/Documentation/networking/ipvs-sysctl.txt @@ -181,6 +181,19 @@ snat_reroute - BOOLEAN always be the same as the original route so it is an optimisation to disable snat_reroute and avoid the recalculation. +sync_persist_mode - INTEGER + default 0 + + Controls the synchronisation of connections when using persistence + + 0: All types of connections are synchronised + 1: Attempt to reduce the synchronisation traffic depending on + the connection type. For persistent services avoid synchronisation + for normal connections, do it only for persistence templates. + In such case, for TCP and SCTP it may need enabling sloppy_tcp and + sloppy_sctp flags on backup servers. For non-persistent services + such optimization is not applied, mode 0 is assumed. + sync_version - INTEGER default 1 diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index e667df1..f0d70f0 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -975,6 +975,7 @@ struct netns_ipvs { int sysctl_snat_reroute; int sysctl_sync_ver; int sysctl_sync_ports; + int sysctl_sync_persist_mode; unsigned long sysctl_sync_qlen_max; int sysctl_sync_sock_size; int sysctl_cache_bypass; @@ -1076,6 +1077,11 @@ static inline int sysctl_sync_ports(struct netns_ipvs *ipvs) return ACCESS_ONCE(ipvs->sysctl_sync_ports); } +static inline int sysctl_sync_persist_mode(struct netns_ipvs *ipvs) +{ + return ipvs->sysctl_sync_persist_mode; +} + static inline unsigned long sysctl_sync_qlen_max(struct netns_ipvs *ipvs) { return ipvs->sysctl_sync_qlen_max; @@ -1139,6 +1145,11 @@ static inline int sysctl_sync_ports(struct netns_ipvs *ipvs) return 1; } +static inline int sysctl_sync_persist_mode(struct netns_ipvs *ipvs) +{ + return 0; +} + static inline unsigned long sysctl_sync_qlen_max(struct netns_ipvs *ipvs) { return IPVS_SYNC_QLEN_MAX; diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index da035fc..c8148e4 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -1715,6 +1715,12 @@ static struct ctl_table vs_vars[] = { .proc_handler = &proc_do_sync_ports, }, { + .procname = "sync_persist_mode", + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + { .procname = "sync_qlen_max", .maxlen = sizeof(unsigned long), .mode = 0644, @@ -3729,6 +3735,7 @@ static int __net_init ip_vs_control_net_init_sysctl(struct net *net) tbl[idx++].data = &ipvs->sysctl_sync_ver; ipvs->sysctl_sync_ports = 1; tbl[idx++].data = &ipvs->sysctl_sync_ports; + tbl[idx++].data = &ipvs->sysctl_sync_persist_mode; ipvs->sysctl_sync_qlen_max = nr_free_buffer_pages() / 32; tbl[idx++].data = &ipvs->sysctl_sync_qlen_max; ipvs->sysctl_sync_sock_size = 0; diff --git a/net/netfilter/ipvs/ip_vs_sync.c b/net/netfilter/ipvs/ip_vs_sync.c index 2fc6639..f448471 100644 --- a/net/netfilter/ipvs/ip_vs_sync.c +++ b/net/netfilter/ipvs/ip_vs_sync.c @@ -425,6 +425,16 @@ ip_vs_sync_buff_create_v0(struct netns_ipvs *ipvs) return sb; } +/* Check if connection is controlled by persistence */ +static inline bool in_persistence(struct ip_vs_conn *cp) +{ + for (cp = cp->control; cp; cp = cp->control) { + if (cp->flags & IP_VS_CONN_F_TEMPLATE) + return true; + } + return false; +} + /* Check if conn should be synced. * pkts: conn packets, use sysctl_sync_threshold to avoid packet check * - (1) sync_refresh_period: reduce sync rate. Additionally, retry @@ -447,6 +457,8 @@ static int ip_vs_sync_conn_needed(struct netns_ipvs *ipvs, /* Check if we sync in current state */ if (unlikely(cp->flags & IP_VS_CONN_F_TEMPLATE)) force = 0; + else if (unlikely(sysctl_sync_persist_mode(ipvs) && in_persistence(cp))) + return 0; else if (likely(cp->protocol == IPPROTO_TCP)) { if (!((1 << cp->state) & ((1 << IP_VS_TCP_S_ESTABLISHED) | |