[RFC net-next 0/1] seg6: Counters for SRv6 Behaviors

[RFC net-next 0/1] seg6: Counters for SRv6 Behaviors

[Andrea Mayer]

This patch provides counters for SRv6 Behaviors as defined in [1],
section 6. For each SRv6 Behavior instance, counters defined in [1] are:

 - the total number of packets that have been correctly processed;
 - the total amount of traffic in bytes of all packets that have been
   correctly processed;

In addition, this patch introduces a new counter that counts the number of
packets that have NOT been properly processed (i.e. errors) by an SRv6
Behavior instance.

Counters are not only interesting for network monitoring purposes (i.e.
counting the number of packets processed by a given behavior) but they also
provide a simple tool for checking whether a behavior instance is working
as we expect or not.
Counters can be useful for troubleshooting misconfigured SRv6 networks.
Indeed, an SRv6 Behavior can silently drop packets for very different
reasons (i.e. wrong SID configuration, interfaces set with SID addresses,
etc) without any notification/message to the user.

Due to the nature of SRv6 networks, diagnostic tools such as ping and
traceroute may be ineffective: paths used for reaching a given router can
be totally different from the ones followed by probe packets. In addition,
paths are often asymmetrical and this makes it even more difficult to keep
up with the journey of the packets and to understand which behaviors are
actually processing our traffic.

When counters are enabled on an SRv6 Behavior instance, it is possible to
verify if packets are actually processed by such behavior and what is the
outcome of the processing. Therefore, the counters for SRv6 Behaviors offer
an non-invasive observability point which can be leveraged for both traffic
monitoring and troubleshooting purposes.

[1] https://www.rfc-editor.org/rfc/rfc8986.html#name-counters

Troubleshooting using SRv6 Behavior counters
--------------------------------------------

Let's make a brief example to see how helpful counters can be for SRv6
networks. Let's consider a node where an SRv6 End Behavior receives an SRv6
packet whose Segment Left (SL) is equal to 0. In this case, the End
Behavior (which accepts only packets with SL >= 1) discards the packet and
increases the error counter.
This information can be leveraged by the network operator for
troubleshooting. Indeed, the error counter is telling the user that the
packet:

  (i) arrived at the node;
 (ii) the packet has been taken into account by the SRv6 End behavior;
(iii) but an error has occurred during the processing.

The error (iii) could be caused by different reasons, such as wrong route
settings on the node or due to an invalid SID List carried by the SRv6
packet. Anyway, the error counter is used to exclude that the packet did
not arrive at the node or it has not been processed by the behavior at
all.

Turning on/off counters for SRv6 Behaviors
------------------------------------------

Each SRv6 Behavior instance can be configured, at the time of its creation,
to make use of counters.
This is done through iproute2 which allows the user to create an SRv6
Behavior instance specifying the optional "count" attribute as shown in the
following example:

 $ ip -6 route add 2001:db8::1 encap seg6local action End count dev eth0

per-behavior counters can be shown by adding "-s" to the iproute2 command
line, i.e.:

 $ ip -s -6 route show 2001:db8::1
 2001:db8::1 encap seg6local action End packets 0 bytes 0 errors 0 dev eth0


####################################################


Impact of counters for SRv6 Behaviors on performance
====================================================

To determine the performance impact due to the introduction of counters in
the SRv6 Behavior subsystem, we have carried out extensive tests.

We chose to test the throughput achieved by the SRv6 End.DX2 Behavior
because, among all the other behaviors implemented so far, it reaches the
highest throughput which is around 1.5 Mpps (per core at 2.4 GHz on a
Xeon(R) CPU E5-2630 v3) on kernel 5.12-rc2 using packets of size ~ 100
bytes.

Three different tests were conducted in order to evaluate the overall
throughput of the SRv6 End.DX2 Behavior in the following scenarios:

 1) vanilla kernel (without the SRv6 Behavior counters patch) and a single
    instance of an SRv6 End.DX2 Behavior;
 2) patched kernel with SRv6 Behavior counters and a single instance of
    an SRv6 End.DX2 Behavior with counters turned off;
 3) patched kernel with SRv6 Behavior counters and a single instance of
    SRv6 End.DX2 Behavior with counters turned on.

All tests were performed on a testbed deployed on the CloudLab facilities
[2], a flexible infrastructure dedicated to scientific research on the
future of Cloud Computing.


Results of tests are shown in the following table:

Scenario (1): average 1504764,81 pps (~1504,76 kpps); std. dev 3956,82 pps
Scenario (2): average 1501469,78 pps (~1501,47 kpps); std. dev 2979,85 pps
Scenario (3): average 1501315,13 pps (~1501,32 kpps); std. dev 2956,00 pps

As can be observed, throughputs achieved in scenarios (2),(3) did not
suffer any observable degradation compared to scenario (1).

Comments, suggestions and improvements are very welcome!

Thanks,
Andrea

[2] https://www.cloudlab.us

Andrea Mayer (1):
  seg6: add counters support for SRv6 Behaviors

 include/uapi/linux/seg6_local.h |   8 ++
 net/ipv6/seg6_local.c           | 133 +++++++++++++++++++++++++++++++-
 2 files changed, 139 insertions(+), 2 deletions(-)

-- 
2.20.1

[RFC net-next 1/1] seg6: add counters support for SRv6 Behaviors

[Andrea Mayer]

This patch provides counters for SRv6 Behaviors as defined in [1], section
6. For each SRv6 Behavior instance, the counters defined in [1] are:

 - the total number of packets that have been correctly processed;
 - the total amount of traffic in bytes of all packets that have been
   correctly processed;

In addition, we introduces a new counter that counts the number of packets
that have NOT been properly processed (i.e. errors) by an SRv6 Behavior
instance.

Each SRv6 Behavior instance can be configured, at the time of its creation,
to make use of counters.
This is done through iproute2 which allows the user to create an SRv6
Behavior instance specifying the optional "count" attribute as shown in the
following example:

 $ ip -6 route add 2001:db8::1 encap seg6local action End count dev eth0

per-behavior counters can be shown by adding "-s" to the iproute2 command
line, i.e.:

 $ ip -s -6 route show 2001:db8::1
 2001:db8::1 encap seg6local action End packets 0 bytes 0 errors 0 dev eth0

[1] https://www.rfc-editor.org/rfc/rfc8986.html#name-counters

Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
---
 include/uapi/linux/seg6_local.h |   8 ++
 net/ipv6/seg6_local.c           | 133 +++++++++++++++++++++++++++++++-
 2 files changed, 139 insertions(+), 2 deletions(-)

diff --git a/include/uapi/linux/seg6_local.h b/include/uapi/linux/seg6_local.h
index 3b39ef1dbb46..ae5e3fd12b73 100644
--- a/include/uapi/linux/seg6_local.h
+++ b/include/uapi/linux/seg6_local.h
@@ -27,6 +27,7 @@ enum {
 	SEG6_LOCAL_OIF,
 	SEG6_LOCAL_BPF,
 	SEG6_LOCAL_VRFTABLE,
+	SEG6_LOCAL_COUNTERS,
 	__SEG6_LOCAL_MAX,
 };
 #define SEG6_LOCAL_MAX (__SEG6_LOCAL_MAX - 1)
@@ -78,4 +79,11 @@ enum {
 
 #define SEG6_LOCAL_BPF_PROG_MAX (__SEG6_LOCAL_BPF_PROG_MAX - 1)
 
+/* SRv6 Behavior counters */
+struct seg6_local_counters {
+	__u64 rx_packets;
+	__u64 rx_bytes;
+	__u64 rx_errors;
+};
+
 #endif
diff --git a/net/ipv6/seg6_local.c b/net/ipv6/seg6_local.c
index 8936f48570fc..0f905a4410bd 100644
--- a/net/ipv6/seg6_local.c
+++ b/net/ipv6/seg6_local.c
@@ -93,6 +93,20 @@ struct seg6_end_dt_info {
 	int hdrlen;
 };
 
+struct pcpu_seg6_local_counters {
+	u64_stats_t rx_packets;
+	u64_stats_t rx_bytes;
+	u64_stats_t rx_errors;
+
+	struct u64_stats_sync syncp;
+};
+
+#define seg6_local_alloc_pcpu_counters(__gfp)				\
+	__netdev_alloc_pcpu_stats(struct pcpu_seg6_local_counters,	\
+				  ((__gfp) | __GFP_ZERO))
+
+#define SEG6_F_LOCAL_COUNTERS	SEG6_F_ATTR(SEG6_LOCAL_COUNTERS)
+
 struct seg6_local_lwt {
 	int action;
 	struct ipv6_sr_hdr *srh;
@@ -105,6 +119,7 @@ struct seg6_local_lwt {
 #ifdef CONFIG_NET_L3_MASTER_DEV
 	struct seg6_end_dt_info dt_info;
 #endif
+	struct pcpu_seg6_local_counters __percpu *pcpu_counters;
 
 	int headroom;
 	struct seg6_action_desc *desc;
@@ -878,36 +893,43 @@ static struct seg6_action_desc seg6_action_table[] = {
 	{
 		.action		= SEG6_LOCAL_ACTION_END,
 		.attrs		= 0,
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_X,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_NH6),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_x,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_T,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_TABLE),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_t,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_DX2,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_OIF),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_dx2,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_DX6,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_NH6),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_dx6,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_DX4,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_NH4),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_dx4,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_DT4,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_VRFTABLE),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 #ifdef CONFIG_NET_L3_MASTER_DEV
 		.input		= input_action_end_dt4,
 		.slwt_ops	= {
@@ -919,30 +941,35 @@ static struct seg6_action_desc seg6_action_table[] = {
 		.action		= SEG6_LOCAL_ACTION_END_DT6,
 #ifdef CONFIG_NET_L3_MASTER_DEV
 		.attrs		= 0,
-		.optattrs	= SEG6_F_ATTR(SEG6_LOCAL_TABLE) |
+		.optattrs	= SEG6_F_LOCAL_COUNTERS		|
+				  SEG6_F_ATTR(SEG6_LOCAL_TABLE) |
 				  SEG6_F_ATTR(SEG6_LOCAL_VRFTABLE),
 		.slwt_ops	= {
 					.build_state = seg6_end_dt6_build,
 				  },
 #else
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_TABLE),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 #endif
 		.input		= input_action_end_dt6,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_B6,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_SRH),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_b6,
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_B6_ENCAP,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_SRH),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_b6_encap,
 		.static_headroom	= sizeof(struct ipv6hdr),
 	},
 	{
 		.action		= SEG6_LOCAL_ACTION_END_BPF,
 		.attrs		= SEG6_F_ATTR(SEG6_LOCAL_BPF),
+		.optattrs	= SEG6_F_LOCAL_COUNTERS,
 		.input		= input_action_end_bpf,
 	},
 
@@ -963,11 +990,36 @@ static struct seg6_action_desc *__get_action_desc(int action)
 	return NULL;
 }
 
+static bool seg6_lwtunnel_counters_enabled(struct seg6_local_lwt *slwt)
+{
+	return slwt->parsed_optattrs & SEG6_F_LOCAL_COUNTERS;
+}
+
+static void seg6_local_update_rx_counters(struct seg6_local_lwt *slwt,
+					  int len, int err)
+{
+	struct pcpu_seg6_local_counters *pcounters;
+
+	pcounters = this_cpu_ptr(slwt->pcpu_counters);
+	u64_stats_update_begin(&pcounters->syncp);
+
+	if (likely(!err)) {
+		u64_stats_inc(&pcounters->rx_packets);
+		u64_stats_add(&pcounters->rx_bytes, len);
+	} else {
+		u64_stats_inc(&pcounters->rx_errors);
+	}
+
+	u64_stats_update_end(&pcounters->syncp);
+}
+
 static int seg6_local_input(struct sk_buff *skb)
 {
 	struct dst_entry *orig_dst = skb_dst(skb);
 	struct seg6_action_desc *desc;
 	struct seg6_local_lwt *slwt;
+	int len = skb->len;
+	int rc;
 
 	if (skb->protocol != htons(ETH_P_IPV6)) {
 		kfree_skb(skb);
@@ -977,7 +1029,14 @@ static int seg6_local_input(struct sk_buff *skb)
 	slwt = seg6_local_lwtunnel(orig_dst->lwtstate);
 	desc = slwt->desc;
 
-	return desc->input(skb, slwt);
+	rc = desc->input(skb, slwt);
+
+	if (!seg6_lwtunnel_counters_enabled(slwt))
+		return rc;
+
+	seg6_local_update_rx_counters(slwt, len, rc);
+
+	return rc;
 }
 
 static const struct nla_policy seg6_local_policy[SEG6_LOCAL_MAX + 1] = {
@@ -992,6 +1051,8 @@ static const struct nla_policy seg6_local_policy[SEG6_LOCAL_MAX + 1] = {
 	[SEG6_LOCAL_IIF]	= { .type = NLA_U32 },
 	[SEG6_LOCAL_OIF]	= { .type = NLA_U32 },
 	[SEG6_LOCAL_BPF]	= { .type = NLA_NESTED },
+	[SEG6_LOCAL_COUNTERS]	= { .type = NLA_BINARY,
+				    .len = sizeof(struct seg6_local_counters) },
 };
 
 static int parse_nla_srh(struct nlattr **attrs, struct seg6_local_lwt *slwt)
@@ -1296,6 +1357,67 @@ static void destroy_attr_bpf(struct seg6_local_lwt *slwt)
 		bpf_prog_put(slwt->bpf.prog);
 }
 
+static int parse_nla_counters(struct nlattr **attrs,
+			      struct seg6_local_lwt *slwt)
+{
+	struct pcpu_seg6_local_counters __percpu *pcounters;
+
+	pcounters = seg6_local_alloc_pcpu_counters(GFP_KERNEL);
+	if (!pcounters)
+		return -ENOMEM;
+
+	slwt->pcpu_counters = pcounters;
+
+	return 0;
+}
+
+static int put_nla_counters(struct sk_buff *skb, struct seg6_local_lwt *slwt)
+{
+	struct seg6_local_counters counters = { 0, 0, 0 };
+	struct nlattr *nla;
+	int i;
+
+	nla = nla_reserve(skb, SEG6_LOCAL_COUNTERS, sizeof(counters));
+	if (!nla)
+		return -EMSGSIZE;
+
+	for_each_possible_cpu(i) {
+		struct pcpu_seg6_local_counters *pcounters;
+		u64 rx_packets, rx_bytes, rx_errors;
+		unsigned int start;
+
+		pcounters = per_cpu_ptr(slwt->pcpu_counters, i);
+		do {
+			start = u64_stats_fetch_begin_irq(&pcounters->syncp);
+
+			rx_packets = u64_stats_read(&pcounters->rx_packets);
+			rx_bytes = u64_stats_read(&pcounters->rx_bytes);
+			rx_errors = u64_stats_read(&pcounters->rx_errors);
+
+		} while (u64_stats_fetch_retry_irq(&pcounters->syncp, start));
+
+		counters.rx_packets += rx_packets;
+		counters.rx_bytes += rx_bytes;
+		counters.rx_errors += rx_errors;
+	}
+
+	memcpy(nla_data(nla), &counters, sizeof(counters));
+
+	return 0;
+}
+
+static int cmp_nla_counters(struct seg6_local_lwt *a, struct seg6_local_lwt *b)
+{
+	/* a and b are equals if both have pcpu_counters set or not */
+	return (!!((unsigned long)a->pcpu_counters)) ^
+		(!!((unsigned long)b->pcpu_counters));
+}
+
+static void destroy_attr_counters(struct seg6_local_lwt *slwt)
+{
+	free_percpu(slwt->pcpu_counters);
+}
+
 struct seg6_action_param {
 	int (*parse)(struct nlattr **attrs, struct seg6_local_lwt *slwt);
 	int (*put)(struct sk_buff *skb, struct seg6_local_lwt *slwt);
@@ -1343,6 +1465,10 @@ static struct seg6_action_param seg6_action_params[SEG6_LOCAL_MAX + 1] = {
 				    .put = put_nla_vrftable,
 				    .cmp = cmp_nla_vrftable },
 
+	[SEG6_LOCAL_COUNTERS]	= { .parse = parse_nla_counters,
+				    .put = put_nla_counters,
+				    .cmp = cmp_nla_counters,
+				    .destroy = destroy_attr_counters },
 };
 
 /* call the destroy() callback (if available) for each set attribute in
@@ -1645,6 +1771,9 @@ static int seg6_local_get_encap_size(struct lwtunnel_state *lwt)
 	if (attrs & SEG6_F_ATTR(SEG6_LOCAL_VRFTABLE))
 		nlsize += nla_total_size(4);
 
+	if (attrs & SEG6_F_LOCAL_COUNTERS)
+		nlsize += nla_total_size(sizeof(struct seg6_local_counters));
+
 	return nlsize;
 }
 
-- 
2.20.1

Re: [RFC net-next 1/1] seg6: add counters support for SRv6 Behaviors

[Jakub Kicinski]

On Wed,  7 Apr 2021 20:03:32 +0200 Andrea Mayer wrote:
> This patch provides counters for SRv6 Behaviors as defined in [1], section
> 6. For each SRv6 Behavior instance, the counters defined in [1] are:
> 
>  - the total number of packets that have been correctly processed;
>  - the total amount of traffic in bytes of all packets that have been
>    correctly processed;
> 
> In addition, we introduces a new counter that counts the number of packets
> that have NOT been properly processed (i.e. errors) by an SRv6 Behavior
> instance.
> 
> Each SRv6 Behavior instance can be configured, at the time of its creation,
> to make use of counters.
> This is done through iproute2 which allows the user to create an SRv6
> Behavior instance specifying the optional "count" attribute as shown in the
> following example:
> 
>  $ ip -6 route add 2001:db8::1 encap seg6local action End count dev eth0
> 
> per-behavior counters can be shown by adding "-s" to the iproute2 command
> line, i.e.:
> 
>  $ ip -s -6 route show 2001:db8::1
>  2001:db8::1 encap seg6local action End packets 0 bytes 0 errors 0 dev eth0
> 
> [1] https://www.rfc-editor.org/rfc/rfc8986.html#name-counters
> 
> Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>

> +static int put_nla_counters(struct sk_buff *skb, struct seg6_local_lwt *slwt)
> +{
> +	struct seg6_local_counters counters = { 0, 0, 0 };
> +	struct nlattr *nla;
> +	int i;
> +
> +	nla = nla_reserve(skb, SEG6_LOCAL_COUNTERS, sizeof(counters));
> +	if (!nla)
> +		return -EMSGSIZE;

nla_reserve_64bit(), IIUC netlink guarantees alignment of 64 bit values.

Re: [RFC net-next 1/1] seg6: add counters support for SRv6 Behaviors

[David Ahern]

On 4/7/21 12:03 PM, Andrea Mayer wrote:
> diff --git a/include/uapi/linux/seg6_local.h b/include/uapi/linux/seg6_local.h
> index 3b39ef1dbb46..ae5e3fd12b73 100644
> --- a/include/uapi/linux/seg6_local.h
> +++ b/include/uapi/linux/seg6_local.h
> @@ -27,6 +27,7 @@ enum {
>  	SEG6_LOCAL_OIF,
>  	SEG6_LOCAL_BPF,
>  	SEG6_LOCAL_VRFTABLE,
> +	SEG6_LOCAL_COUNTERS,
>  	__SEG6_LOCAL_MAX,
>  };
>  #define SEG6_LOCAL_MAX (__SEG6_LOCAL_MAX - 1)
> @@ -78,4 +79,11 @@ enum {
>  
>  #define SEG6_LOCAL_BPF_PROG_MAX (__SEG6_LOCAL_BPF_PROG_MAX - 1)
>  
> +/* SRv6 Behavior counters */
> +struct seg6_local_counters {
> +	__u64 rx_packets;
> +	__u64 rx_bytes;
> +	__u64 rx_errors;
> +};
> +
>  #endif

It's highly likely that more stats would get added over time. It would
be good to document that here for interested parties and then make sure
iproute2 can handle different sized stats structs. e.g., commit support
to your repo, then add a new one (e.g, rx_drops) and verify the
combinations handle it. e.g., old kernel - new iproute2, new kernel -
old iproute, old - old and new-new.

Re: [RFC net-next 0/1] seg6: Counters for SRv6 Behaviors

[David Ahern]

Since this is a single patch set, just put this good cover letter
content as the message in the patch.

Re: [RFC net-next 1/1] seg6: add counters support for SRv6 Behaviors

[Andrea Mayer]

On Wed, 7 Apr 2021 13:24:04 -0700
Jakub Kicinski <kuba@kernel.org> wrote:

> On Wed,  7 Apr 2021 20:03:32 +0200 Andrea Mayer wrote:
> > This patch provides counters for SRv6 Behaviors as defined in [1], section
> > 6. For each SRv6 Behavior instance, the counters defined in [1] are:
> > 
> >  - the total number of packets that have been correctly processed;
> >  - the total amount of traffic in bytes of all packets that have been
> >    correctly processed;
> > 
> > In addition, we introduces a new counter that counts the number of packets
> > that have NOT been properly processed (i.e. errors) by an SRv6 Behavior
> > instance.
> > 
> > Each SRv6 Behavior instance can be configured, at the time of its creation,
> > to make use of counters.
> > This is done through iproute2 which allows the user to create an SRv6
> > Behavior instance specifying the optional "count" attribute as shown in the
> > following example:
> > 
> >  $ ip -6 route add 2001:db8::1 encap seg6local action End count dev eth0
> > 
> > per-behavior counters can be shown by adding "-s" to the iproute2 command
> > line, i.e.:
> > 
> >  $ ip -s -6 route show 2001:db8::1
> >  2001:db8::1 encap seg6local action End packets 0 bytes 0 errors 0 dev eth0
> > 
> > [1] https://www.rfc-editor.org/rfc/rfc8986.html#name-counters
> > 
> > Signed-off-by: Andrea Mayer <andrea.mayer@uniroma2.it>
> 
> > +static int put_nla_counters(struct sk_buff *skb, struct seg6_local_lwt *slwt)
> > +{
> > +	struct seg6_local_counters counters = { 0, 0, 0 };
> > +	struct nlattr *nla;
> > +	int i;
> > +
> > +	nla = nla_reserve(skb, SEG6_LOCAL_COUNTERS, sizeof(counters));
> > +	if (!nla)
> > +		return -EMSGSIZE;
> 
> nla_reserve_64bit(), IIUC netlink guarantees alignment of 64 bit values.

Hi Jakub, thanks for your review!

Yes, we should guarantee alignment of 64 bit values.
I will definitely follow your advice.

Andrea

Re: [RFC net-next 1/1] seg6: add counters support for SRv6 Behaviors

[Andrea Mayer]

On Wed, 7 Apr 2021 16:55:41 -0600
David Ahern <dsahern@gmail.com> wrote:

> On 4/7/21 12:03 PM, Andrea Mayer wrote:
> > diff --git a/include/uapi/linux/seg6_local.h b/include/uapi/linux/seg6_local.h
> > index 3b39ef1dbb46..ae5e3fd12b73 100644
> > --- a/include/uapi/linux/seg6_local.h
> > +++ b/include/uapi/linux/seg6_local.h
> > @@ -27,6 +27,7 @@ enum {
> >  	SEG6_LOCAL_OIF,
> >  	SEG6_LOCAL_BPF,
> >  	SEG6_LOCAL_VRFTABLE,
> > +	SEG6_LOCAL_COUNTERS,
> >  	__SEG6_LOCAL_MAX,
> >  };
> >  #define SEG6_LOCAL_MAX (__SEG6_LOCAL_MAX - 1)
> > @@ -78,4 +79,11 @@ enum {
> >  
> >  #define SEG6_LOCAL_BPF_PROG_MAX (__SEG6_LOCAL_BPF_PROG_MAX - 1)
> >  
> > +/* SRv6 Behavior counters */
> > +struct seg6_local_counters {
> > +	__u64 rx_packets;
> > +	__u64 rx_bytes;
> > +	__u64 rx_errors;
> > +};
> > +
> >  #endif
> 
> It's highly likely that more stats would get added over time. It would
> be good to document that here for interested parties and then make sure
> iproute2 can handle different sized stats structs. e.g., commit support
> to your repo, then add a new one (e.g, rx_drops) and verify the
> combinations handle it. e.g., old kernel - new iproute2, new kernel -
> old iproute, old - old and new-new.
> 

Hi David,
thanks for your review.

I totally agree with you: we may want to add other counters in the future, even
if they are not considered in RFC8986.

With that in mind, the shared struct seg6_local_counters is not the best way to
go if we want to add other counters (because it will be difficult to manage
different sized structures when considering different kernel/iproute2 versions).

To make it easier adding new counters, instead of sharing the struct
seg6_local_counters, I would use netlink nested attributes to exchange counters
individually. In this way, only recognized (nested) attributes can be processed
by both the kernel and iproute2.

For example:

enum {
       SEG6_LOCAL_CNT_UNSPEC,
       SEG6_LOCAL_CNT_PAD,             /* padding for 64 bits values */
       SEG6_LOCAL_CNT_RX_PACKETS,
       SEG6_LOCAL_CNT_RX_BYTES,
       SEG6_LOCAL_CNT_RX_ERRORS,
       __SEG6_LOCAL_CNT_MAX,
};
#define SEG6_LOCAL_CNT_MAX (__SEG6_LOCAL_CNT_MAX - 1)

updating the policy for SEG6_LOCAL_COUNTERS to NLA_NESTED.

Then, I create a new policy for counters which handles each supported
counter separately.

static const struct
nla_policy seg6_local_counters_policy[SEG6_LOCAL_CNT_MAX + 1] = {
       [SEG6_LOCAL_CNT_RX_PACKETS]     = { .type = NLA_U64 },
       [SEG6_LOCAL_CNT_RX_BYTES]       = { .type = NLA_U64 },
       [SEG6_LOCAL_CNT_RX_ERRORS]      = { .type = NLA_U64 },
};

At the end, I update the parse_nla_counters(), put_nla_counters(), etc
according to the changes, i.e:
 - nla_parse_nested() in parse_nla_counters();
 - nla_nest_{start/end}() and for each supported counter nla_put_u64_64bit()
   in put_nla_counters().

On the iproute2 side, we have to update the code to reflect the changes
discussed above. 

I plan to issue an RFC v2 in a few days.

Andrea