From: kaber@trash.net
To: davem@davemloft.net
Cc: netfilter-devel@vger.kernel.org, netdev@vger.kernel.org
Subject: [PATCH 2/9] IPVS: one-packet scheduling
Date: Fri, 2 Jul 2010 11:52:15 +0200 [thread overview]
Message-ID: <1278064342-19059-3-git-send-email-kaber@trash.net> (raw)
In-Reply-To: <1278064342-19059-1-git-send-email-kaber@trash.net>
From: Nick Chalk <nick@loadbalancer.org>
Allow one-packet scheduling for UDP connections. When the fwmark-based or
normal virtual service is marked with '-o' or '--ops' options all
connections are created only to schedule one packet. Useful to schedule UDP
packets from same client port to different real servers. Recommended with
RR or WRR schedulers (the connections are not visible with ipvsadm -L).
Signed-off-by: Nick Chalk <nick@loadbalancer.org>
Signed-off-by: Simon Horman <horms@verge.net.au>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/linux/ip_vs.h | 2 ++
net/netfilter/ipvs/ip_vs_conn.c | 10 +++++++---
net/netfilter/ipvs/ip_vs_core.c | 20 ++++++++++++++++----
net/netfilter/ipvs/ip_vs_ctl.c | 10 ++++++----
4 files changed, 31 insertions(+), 11 deletions(-)
diff --git a/include/linux/ip_vs.h b/include/linux/ip_vs.h
index dfc1703..9708de2 100644
--- a/include/linux/ip_vs.h
+++ b/include/linux/ip_vs.h
@@ -19,6 +19,7 @@
*/
#define IP_VS_SVC_F_PERSISTENT 0x0001 /* persistent port */
#define IP_VS_SVC_F_HASHED 0x0002 /* hashed entry */
+#define IP_VS_SVC_F_ONEPACKET 0x0004 /* one-packet scheduling */
/*
* Destination Server Flags
@@ -85,6 +86,7 @@
#define IP_VS_CONN_F_SEQ_MASK 0x0600 /* in/out sequence mask */
#define IP_VS_CONN_F_NO_CPORT 0x0800 /* no client port set yet */
#define IP_VS_CONN_F_TEMPLATE 0x1000 /* template, not connection */
+#define IP_VS_CONN_F_ONE_PACKET 0x2000 /* forward only one packet */
#define IP_VS_SCHEDNAME_MAXLEN 16
#define IP_VS_IFNAME_MAXLEN 16
diff --git a/net/netfilter/ipvs/ip_vs_conn.c b/net/netfilter/ipvs/ip_vs_conn.c
index d8f7e8e..717e623 100644
--- a/net/netfilter/ipvs/ip_vs_conn.c
+++ b/net/netfilter/ipvs/ip_vs_conn.c
@@ -158,6 +158,9 @@ static inline int ip_vs_conn_hash(struct ip_vs_conn *cp)
unsigned hash;
int ret;
+ if (cp->flags & IP_VS_CONN_F_ONE_PACKET)
+ return 0;
+
/* Hash by protocol, client address and port */
hash = ip_vs_conn_hashkey(cp->af, cp->protocol, &cp->caddr, cp->cport);
@@ -355,8 +358,9 @@ struct ip_vs_conn *ip_vs_conn_out_get
*/
void ip_vs_conn_put(struct ip_vs_conn *cp)
{
- /* reset it expire in its timeout */
- mod_timer(&cp->timer, jiffies+cp->timeout);
+ unsigned long t = (cp->flags & IP_VS_CONN_F_ONE_PACKET) ?
+ 0 : cp->timeout;
+ mod_timer(&cp->timer, jiffies+t);
__ip_vs_conn_put(cp);
}
@@ -649,7 +653,7 @@ static void ip_vs_conn_expire(unsigned long data)
/*
* unhash it if it is hashed in the conn table
*/
- if (!ip_vs_conn_unhash(cp))
+ if (!ip_vs_conn_unhash(cp) && !(cp->flags & IP_VS_CONN_F_ONE_PACKET))
goto expire_later;
/*
diff --git a/net/netfilter/ipvs/ip_vs_core.c b/net/netfilter/ipvs/ip_vs_core.c
index 1cd6e3f..50907d8 100644
--- a/net/netfilter/ipvs/ip_vs_core.c
+++ b/net/netfilter/ipvs/ip_vs_core.c
@@ -194,6 +194,7 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
struct ip_vs_dest *dest;
struct ip_vs_conn *ct;
__be16 dport; /* destination port to forward */
+ __be16 flags;
union nf_inet_addr snet; /* source network of the client,
after masking */
@@ -340,6 +341,10 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
dport = ports[1];
}
+ flags = (svc->flags & IP_VS_SVC_F_ONEPACKET
+ && iph.protocol == IPPROTO_UDP)?
+ IP_VS_CONN_F_ONE_PACKET : 0;
+
/*
* Create a new connection according to the template
*/
@@ -347,7 +352,7 @@ ip_vs_sched_persist(struct ip_vs_service *svc,
&iph.saddr, ports[0],
&iph.daddr, ports[1],
&dest->addr, dport,
- 0,
+ flags,
dest);
if (cp == NULL) {
ip_vs_conn_put(ct);
@@ -377,7 +382,7 @@ ip_vs_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
struct ip_vs_conn *cp = NULL;
struct ip_vs_iphdr iph;
struct ip_vs_dest *dest;
- __be16 _ports[2], *pptr;
+ __be16 _ports[2], *pptr, flags;
ip_vs_fill_iphdr(svc->af, skb_network_header(skb), &iph);
pptr = skb_header_pointer(skb, iph.len, sizeof(_ports), _ports);
@@ -407,6 +412,10 @@ ip_vs_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
return NULL;
}
+ flags = (svc->flags & IP_VS_SVC_F_ONEPACKET
+ && iph.protocol == IPPROTO_UDP)?
+ IP_VS_CONN_F_ONE_PACKET : 0;
+
/*
* Create a connection entry.
*/
@@ -414,7 +423,7 @@ ip_vs_schedule(struct ip_vs_service *svc, const struct sk_buff *skb)
&iph.saddr, pptr[0],
&iph.daddr, pptr[1],
&dest->addr, dest->port ? dest->port : pptr[1],
- 0,
+ flags,
dest);
if (cp == NULL)
return NULL;
@@ -464,6 +473,9 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
if (sysctl_ip_vs_cache_bypass && svc->fwmark && unicast) {
int ret, cs;
struct ip_vs_conn *cp;
+ __u16 flags = (svc->flags & IP_VS_SVC_F_ONEPACKET &&
+ iph.protocol == IPPROTO_UDP)?
+ IP_VS_CONN_F_ONE_PACKET : 0;
union nf_inet_addr daddr = { .all = { 0, 0, 0, 0 } };
ip_vs_service_put(svc);
@@ -474,7 +486,7 @@ int ip_vs_leave(struct ip_vs_service *svc, struct sk_buff *skb,
&iph.saddr, pptr[0],
&iph.daddr, pptr[1],
&daddr, 0,
- IP_VS_CONN_F_BYPASS,
+ IP_VS_CONN_F_BYPASS | flags,
NULL);
if (cp == NULL)
return NF_DROP;
diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c
index 36dc1d8..0f0c079 100644
--- a/net/netfilter/ipvs/ip_vs_ctl.c
+++ b/net/netfilter/ipvs/ip_vs_ctl.c
@@ -1864,14 +1864,16 @@ static int ip_vs_info_seq_show(struct seq_file *seq, void *v)
svc->scheduler->name);
else
#endif
- seq_printf(seq, "%s %08X:%04X %s ",
+ seq_printf(seq, "%s %08X:%04X %s %s ",
ip_vs_proto_name(svc->protocol),
ntohl(svc->addr.ip),
ntohs(svc->port),
- svc->scheduler->name);
+ svc->scheduler->name,
+ (svc->flags & IP_VS_SVC_F_ONEPACKET)?"ops ":"");
} else {
- seq_printf(seq, "FWM %08X %s ",
- svc->fwmark, svc->scheduler->name);
+ seq_printf(seq, "FWM %08X %s %s",
+ svc->fwmark, svc->scheduler->name,
+ (svc->flags & IP_VS_SVC_F_ONEPACKET)?"ops ":"");
}
if (svc->flags & IP_VS_SVC_F_PERSISTENT)
--
1.7.0.4
next prev parent reply other threads:[~2010-07-02 9:52 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-07-02 9:52 [PATCH 0/9] netfilter: netfilter update kaber
2010-07-02 9:52 ` [PATCH 1/9] netfilter: nf_nat: support user-specified SNAT rules in LOCAL_IN kaber
2010-07-02 10:14 ` Jan Engelhardt
2010-07-02 10:17 ` Patrick McHardy
2010-07-02 12:17 ` Jan Engelhardt
2010-07-02 12:35 ` Patrick McHardy
2010-07-02 12:58 ` Jan Engelhardt
2010-07-02 14:07 ` Patrick McHardy
2010-07-02 9:52 ` kaber [this message]
2010-07-02 9:52 ` [PATCH 3/9] netfilter: xt_IDLETIMER needs kdev_t.h kaber
2010-07-02 9:52 ` [PATCH 4/9] netfilter: fix simple typo in KConfig for netfiltert xt_TEE kaber
2010-07-02 9:52 ` [PATCH 5/9] netfilter: xt_connbytes: Force CT accounting to be enabled kaber
2010-07-02 9:52 ` [PATCH 6/9] netfilter: complete the deprecation of CONFIG_NF_CT_ACCT kaber
2010-07-02 9:52 ` [PATCH 7/9] netfilter: ipt_LOG/ip6t_LOG: remove comparison within loop kaber
2010-07-02 9:52 ` [PATCH 8/9] netfilter: ipt_LOG/ip6t_LOG: add option to print decoded MAC header kaber
2010-07-02 9:52 ` [PATCH 9/9] bridge: add per bridge device controls for invoking iptables kaber
2010-07-03 5:04 ` [PATCH 0/9] netfilter: netfilter update David Miller
2010-07-03 5:44 ` David Miller
2010-07-03 9:06 ` Patrick McHardy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1278064342-19059-3-git-send-email-kaber@trash.net \
--to=kaber@trash.net \
--cc=davem@davemloft.net \
--cc=netdev@vger.kernel.org \
--cc=netfilter-devel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).