Guidelines and Limitations
Currently, the ASIC on the switch forwards the flows only to the default VRF. There is no support now for having collectors in an area other than the default VRF.
Cisco Nexus 3400-S Series NX-OS Programmability Guide, Release 9.3(x)
Bias-Free Language
The documentation set for this product strives to use bias-free language. For the purposes of this documentation set, bias-free is defined as language that does not imply discrimination based on age, disability, gender, racial identity, ethnic identity, sexual orientation, socioeconomic status, and intersectionality. Exceptions may be present in the documentation due to language that is hardcoded in the user interfaces of the product software, language used based on RFP documentation, or language that is used by a referenced third-party product. Learn more about how Cisco is using Inclusive Language.
- Updated:
- March 16, 2020
Chapter: Hardware Telemetry
- Hardware Telemetry
- Guidelines and Limitations
- Telemetry RPM Installation Example
- About Streaming Statistics Export
- SSX Supported Types
- SSX gRPC Filter Types
- SSX Examples
- About Port Counters
- Port Counter Exports
- Port Counter Sample Output
- Guidelines and Limitations for Buffer Drop Capture and Buffer Latency Capture
- Configuring Software Telemetry
- About Buffer Drop Capture
- Buffer Drop Filter Types
- Configuring Buffer Drop Capture
- About Buffer Latency Capture
- Buffer Latency Filter Types
- Configuring Buffer Latency Capture
- TCAM Carving for Flow of Interest
Hardware Telemetry
This chapter contains the following topics:
Telemetry RPM Installation Example
The following is an example of installing the telemetry RPM.
switch# configure terminal
Enter configuration commands, one per line. End with CNTL/Z.
switch(config)# feature bash-shell
switch(config)# end
switch# run bash sudo su
bash-4.3# service docker start >> Starting docker service
Free bootflash: 86936 MB, total bootflash: 116590 MB
Carving docker bootflash storage: 2000 MB
2000+0 records in
2000+0 records out
2000000000 bytes (2.0 GB) copied, 16.5267 s, 121 MB/s
mke2fs 1.42.9 (28-Dec-2013)
fs_types for mke2fs.conf resolution: 'ext4'
Discarding device blocks: done
Discard succeeded and will return 0s - skipping inode table wipe
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
Stride=0 blocks, Stripe width=0 blocks
122160 inodes, 488281 blocks
4882 blocks (1.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=503316480
15 block groups
32768 blocks per group, 32768 fragments per group
8144 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912
Allocating group tables: done
Writing inode tables: done
Creating journal (8192 blocks): done
Writing superblocks and filesystem accounting information: done
tune2fs 1.42.9 (28-Dec-2013)
Filesystem volume name: <none>
Last mounted on: <not available>
Filesystem UUID: 7c841fbf-1972-4082-9ea0-37b77844be3b
Filesystem magic number: 0xEF53
Filesystem revision #: 1 (dynamic)
Filesystem features: has_journal ext_attr resize_inode dir_index filetype extent flex_bg sparse_super large_file huge_file uninit_bg dir_nlink extra_isize
Filesystem flags: signed_directory_hash
Default mount options: user_xattr acl
Filesystem state: clean
Errors behavior: Continue
Filesystem OS type: Linux
Inode count: 122160
Block count: 488281
Reserved block count: 4882
Free blocks: 471692
Free inodes: 122149
First block: 0
Block size: 4096
Fragment size: 4096
Reserved GDT blocks: 119
Blocks per group: 32768
Fragments per group: 32768
Inodes per group: 8144
Inode blocks per group: 509
Flex block group size: 16
Filesystem created: Tue Dec 10 18:46:46 2019
Last mount time: n/a
Last write time: Tue Dec 10 18:46:47 2019
Mount count: 0
Maximum mount count: -1
Last checked: Tue Dec 10 18:46:46 2019
Check interval: 0 (<none>)
Lifetime writes: 33 MB
Reserved blocks uid: 0 (user root)
Reserved blocks gid: 0 (group root)
First inode: 11
Inode size: 256
Required extra isize: 28
Desired extra isize: 28
Journal inode: 8
Default directory hash: half_md4
Directory Hash Seed: bbea6663-7b8a-47e6-8a10-ac87ca23dfcc
Journal backup: inode blocks
Updating certificates in /etc/ssl/certs...
0 added, 0 removed; done.
Running hooks in /etc/ca-certificates/update.d...
done.
Starting dockerd with args '--debug=true': .
bash-4.3# cd /bootflash/
bash-4.3# ls sw_telemetry-2.0-2.x86_64.rpm
sw_telemetry-2.0-2.x86_64.rpm
bash-4.3# rpm -Uvh sw_telemetry-2.0-2.x86_64.rpm >> Installing container RPM
Preparing... ########################################### [100%]
1:sw_telemetry ########################################### [100%]
9da9fb5b1b21: Loading layer [==================================================>] 197.4 MB/197.4 MB
855a0ab0dad9: Loading layer [==================================================>] 13.54 MB/13.54 MB
0fbbc3806bd9: Loading layer [==================================================>] 539.1 kB/539.1 kB
17518e2561b2: Loading layer [==================================================>] 81.41 kB/81.41 kB
a00c7083c0c8: Loading layer [==================================================>] 469.5 kB/469.5 kB
b5b4081bee95: Loading layer [==================================================>] 4.608 kB/4.608 kB
83af9595fe45: Loading layer [==================================================>] 1.536 kB/1.536 kB
96b5c176e3d9: Loading layer [==================================================>] 25.73 MB/25.73 MB
Loaded image: sw_telemetry:latest
bash-4.3# docker images
REPOSITORY TAG IMAGE ID CREATED SIZE
sw_telemetry latest b8f86559a869 36 hours ago 228 MB
bash-4.3# pwd
/bootflash
bash-4.3# cat env.list >> create env.list file with local loopback ip which is reachable to external collector
transport_source_ip=14.13.12.11
bash-4.3# exit
exit
switch# show running-config interface loopback 10
!Command: show running-config interface loopback10
!Running configuration last done at: Tue Dec 10 18:55:59 2019
!Time: Tue Dec 10 18:57:25 2019
version 9.3(3) Bios:version 05.39
interface loopback10
ip address 14.13.12.11/32
ip router ospf 10 area 0.0.0.0
switch# ping 5.1.1.1 source-interface loopback 10 >> 5.1.1.1 is my external collector
PING 5.1.1.1 (5.1.1.1): 56 data bytes
64 bytes from 5.1.1.1: icmp_seq=0 ttl=62 time=0.766 ms
64 bytes from 5.1.1.1: icmp_seq=1 ttl=62 time=0.64 ms
64 bytes from 5.1.1.1: icmp_seq=2 ttl=62 time=0.557 ms
64 bytes from 5.1.1.1: icmp_seq=3 ttl=62 time=0.553 ms
64 bytes from 5.1.1.1: icmp_seq=4 ttl=62 time=0.532 ms
--- 5.1.1.1 ping statistics ---
5 packets transmitted, 5 packets received, 0.00% packet loss
round-trip min/avg/max = 0.532/0.609/0.766 ms
switch# run bash sudo su
bash-4.3# cd /bootflash/
bash-4.3# chkconfig --add docker >> make docker persistent on realod
bash-4.3# docker run -v /var/run/netns:/var/run/netns:ro,rslave --network host --cap-add SYS_ADMIN -it --env-file ./env.list -v /bootflash:/bootflash -p 50051:50051 -p 50052:50052 -d -v /tmp/nginx_local/:/tmp/nginx_local/ -v /etc/localtime:/etc/localtime:ro -v /etc:/etc -v /nxos/tmp:/nxos/tmp --restart unless-stopped sw_telemetry >> starting container
57a65cbceb599f3750c76de1d88749054d8a396c23c1c34e2bd22743a14dbc3e
bash-4.3# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
57a65cbceb59 sw_telemetry "ip netns exec def..." 2 seconds ago Up 1 second determined_meninsky
bash-4.3#
bash-4.3# exit
About Streaming Statistics Export
The Streaming Statistics Export (SSX) module reads statistics from the ASIC and sends them to a remote server (collector) for analysis. SSX can read a register or (directly accessible) memory of the switch.
The contents of registers and the packet format are specified by an instruction sequence.
SSX Supported Types
|
Types |
Description |
|---|---|
|
egress buffer depth |
Streams the buffer occupancy per slice |
|
egress queue depth |
Instant buffer utilization per output queue |
|
egress queue drop |
Tail drops per output queue |
|
ingress queue depth |
Instant buffer utilization per input queue |
|
ingress queue drop |
Overflow drops per input queue |
|
ECN stats |
ECN marked packets per output queue |
SSX gRPC Filter Types
The docker container can be configured using gRPC and the filters that are supported for SSX in the following table. The gRPC message number is the enum number of the filters in the proto file.
|
Filter Types |
Filter_types gRPC Message Number |
|---|---|
|
FILTER_TYPE_EGRESS_BUFFER_DEPTH |
6 |
|
FILTER_TYPE_EGRESS_Q_DEPTH, |
3 |
|
FILTER_TYPE_EGRESS_Q_DROP, |
4 |
|
FILTER_TYPE_INGRESS_Q_DEPTH, |
2 |
|
FILTER_TYPE_INGRESS_Q_DROP, |
4 |
|
FILTER_TYPE_ECN_STATS |
7 |
SSX Examples
SSX Egress Queue Drop
"egressQueueDrop": [
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 0,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 1,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 2,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 3,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 4,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 5,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 6,
"ts": "1562008724381260"
},
{
"QDropBytes": "0",
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 7,
"ts": "1562008724381260"
},
SSX Ingress Queue Occupancy
"ingressQueueOccupancy": [
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 0,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 1,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 2,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 3,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 4,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 5,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 6,
"ts": "1562640434373522"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 7,
"ts": "1562640434373522"
},
SSX Egress Queue Occupancy
"egressQueueOccupancy": [
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 0,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 1,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 2,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 3,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 4,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 5,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 6,
"ts": "1562640434380646"
},
{
"QOcc": "0",
"intf": "Ethernet1/1",
"queueIndex": 7,
"ts": "1562640434380646"
},
SSX Ingress Pause Drops
"data": {
"ingressQueuePauseDrop": [
{
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 4,
"ts": "1562970319568955"
},
{
"QDropPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 3,
"ts": "1562970319568955"
},
SSX Egress Mem Occupancy (per slice)
"egressMemOccupancy": [
{
"EgBuf": "0",
"slice": "0",
"ts": "1573329731594964"
},
{
"EgBuf": "8066",
"slice": "1",
"ts": "1573329731594964"
},
{
"EgBuf": "0",
"slice": "2",
"ts": "1573329731594964"
},
{
"EgBuf": "0",
"slice": "3",
"ts": "1573329731594964"
}
],
SSX ECN
"ecnStats": [
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 0,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 1,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 2,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 3,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 4,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 5,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 6,
"ts": "1562018638423871"
},
{
"ECNBytes": "0",
"ECNPkts": "0",
"intf": "Ethernet1/1",
"queueIndex": 7,
"ts": "1562018638423871"
},
About Port Counters
The following port counters are supported:
|
Port Counter Exports |
Description |
|---|---|
|
Port-in/out stats |
Port input and output stats in pkts/bytes |
|
Port-in/out Discard stats |
Port input and output discard stats in pkts/bytes |
|
Port-in/out Error stats |
Port input and output error stats in pkts/bytes |
|
Port-in/out IPv4 stats |
Port input and output IPV4 stats in pkts/bytes |
|
Port-in/out IPv6 stats |
Port input and output IPV6 stats in pkts/bytes |
|
Port-QoS Group stats |
Per port, per queue egress queue stats in pkts/bytes |
|
Port-PFC stats (TX/RX) |
Per port, per queue PFC stats (TX/RX) |
Port Counter Exports
|
Filter Type |
Filter_types gRPC Message Number |
|---|---|
|
FILTER_TYPE_PORT_COUNTERS |
1 |
Port Counter Sample Output
The following is a sample of port counter output.
"counters": [
{
"interfaceName": "Ethernet1/1",
"portIn": "21900548",
"portInDiscard": "0",
"portInError": "0",
"portInIpv4": "5468000",
"portInIpv4Rate": "0",
"portInIpv6": "5468094",
"portInIpv6Rate": "0",
"portOut": "22919963",
"portOutDiscard": "0",
"portOutError": "0",
"portOutIpv4": "5468000",
"portOutIpv4Rate": "0",
"portOutIpv6": "5468000",
"portOutIpv6Rate": "0",
"portPacketIn": "109722",
"portPacketInIpv4": "27340",
"portPacketInIpv4Rate": "0",
"portPacketInIpv6": "27341",
"portPacketInIpv6Rate": "0",
"portPacketOut": "112917",
"portPacketOutIpv4": "27340",
"portPacketOutIpv4Rate": "0",
"portPacketOutIpv6": "27340",
"portPacketOutIpv6Rate": "0",
"portQosGroup0MulticastTxBytes": "0",
"portQosGroup0MulticastTxPkts": "0",
"portQosGroup0UnicastTxBytes": "440800",
"portQosGroup0UnicastTxPkts": "2204",
"portQosGroup1MulticastTxBytes": "0",
"portQosGroup1MulticastTxPkts": "0",
"portQosGroup1UnicastTxBytes": "21231200",
"portQosGroup1UnicastTxPkts": "106156",
"portQosGroup2MulticastTxBytes": "0",
"portQosGroup2MulticastTxPkts": "0",
"portQosGroup2UnicastTxBytes": "0",
"portQosGroup2UnicastTxPkts": "0",
"portQosGroup3MulticastTxBytes": "0",
"portQosGroup3MulticastTxPkts": "0",
"portQosGroup3UnicastTxBytes": "0",
"portQosGroup3UnicastTxPkts": "0",
"portQosGroup4MulticastTxBytes": "0",
"portQosGroup4MulticastTxPkts": "0",
"portQosGroup4UnicastTxBytes": "0",
"portQosGroup4UnicastTxPkts": "0",
"portQosGroup5MulticastTxBytes": "0",
"portQosGroup5MulticastTxPkts": "0",
"portQosGroup5UnicastTxBytes": "0",
"portQosGroup5UnicastTxPkts": "0",
"portQosGroup6MulticastTxBytes": "0",
"portQosGroup6MulticastTxPkts": "0",
"portQosGroup6UnicastTxBytes": "0",
"portQosGroup6UnicastTxPkts": "0",
"portQosGroup7MulticastTxBytes": "0",
"portQosGroup7MulticastTxPkts": "0",
"portQosGroup7UnicastTxBytes": "0",
"portRx0Pfc": "0",
"portRx1Pfc": "0",
"portRx2Pfc": "0",
"portRx3Pfc": "0",
"portRx4Pfc": "0",
"portRx5Pfc": "0",
"portRx6Pfc": "0",
"portRx7Pfc": "0",
"portTx0Pfc": "0",
"portTx1Pfc": "0",
"portTx2Pfc": "0",
"portTx3Pfc": "0",
"portTx4Pfc": "0",
"portTx5Pfc": "0",
"portTx6Pfc": "0",
"portTx7Pfc": "0",
"timestamp": "1561055719695692554"
},
Guidelines and Limitations for Buffer Drop Capture and Buffer Latency Capture
The following are the Guidelines and Limitations for Buffer Drop Capture (BDC) and Buffer Latency Capture (BLC).
-
These features are supported on the following platforms:
-
Cisco Nexus 3408-S
-
Cisco Nexus 34328D-S
-
Cisco Nexus 34200YC-SM
-
-
Flow of Interest (FoI): You must apply IPv4 and IPv6 ACLs. If you add only an IPv4 ACL, an explicit deny for IPv6 is configured.
-
FoI: Use the same ACL for BDC and BLC.
-
BDC does not work on IB4 and IB5.
-
BDC does not work on no-drop queue.
-
ODM merge does not work for BDC/BLC FoI. By default, statistic-per-entry is present for these ACLs.
-
The hw-telemetry table does not have enough match bits available for ether type (16 bits) to differentiate between IPv4 and IPv6.
-
If the original data traffic and SPAN copy of that traffic is going out on the same IB and if the original data traffic is dropped on a queue where BDC is enabled, you will not see any BDC packets for this traffic.
-
If the BLC port of interest is also configured as SPAN TX source (egress mirror enable), only SPAN destination will receive a copy and the BLC copy is not created. Hence, BLC does not work on this port.
-
We support 1000 PPS for BDC/BLC together.
-
You need to carve the TCAM before using FoI. An error is not displayed when applying FoI without TCAM carving.
Configuring Software Telemetry
Configure software telemetry to support streaming data to the docker container running on the switch. Software telemetry must run on same VRF as the container. For example, if the docker container is running on the default VRF, software telemetry must be configured with the use-vrf default command for streaming the data.
Procedure
| Command or Action | Purpose | |
|---|---|---|
| Step 1 |
configure terminal Example: |
Enter global configuration mode. |
| Step 2 |
feature telemetry Example: |
Enter configuration mode for streaming telemetry. |
| Step 3 |
feature hardware-telemetry Example: |
Enter configuration mode for streaming telemetry. |
| Step 4 |
telemetry Example: |
Enter configuration mode for streaming telemetry. |
| Step 5 |
destination-profile Example: |
Enter the destination-profile command to specify the default destination profile. |
| Step 6 |
use-vrf default Example: |
To specify the destination VRF. |
About Buffer Drop Capture
Buffer Drop Capture (BDC) allows you to capture information about packets which are dropped due to buffer limits. When a packet is dropped due to buffer limits, the device adds metadata to the dropped packets and sends the first 150 bytes of the packet along with metadata to the collector as a BDC packet. The device provides sampling mechanisms such as a probabilistic capture and microburst capture that can be applied to the BDC packets.
BDC allows users to capture information about packets which are dropped due to buffer limits
The dropped packet information can be sent to a remote collector or local CPU.
When a packet is dropped due to buffer limit, metadata (shim header) is added to the dropped packet. The first 150 bytes of the packet along with the metadata is sent to the collector.
The metadata has more details about the dropped packet, the format of metadata is listed below.
The device provides sampling mechanisms such as probabilistic capture and microburst capture which can be applied to the BDC packets.
BDC can be enabled/disabled per Class of Service queue id
Buffer Drop Filter Types
|
Filter Types |
Filter_types gRPC Message Number |
|---|---|
|
FILTER_TYPE_BUFFER_DROP |
8 |
Configuring Buffer Drop Capture
Buffer Drop Capture allows from the capture of information about packets which are dropped due to buffer limits.
Procedure
| Command or Action | Purpose | |||
|---|---|---|---|---|
| Step 1 |
configure terminal Example: |
Enter global configuration mode. |
||
| Step 2 |
hardware-telemetry buffer-drop Example: |
Enable buffer drop configuration. |
||
| Step 3 |
buffer-drop collector name Example: |
Define a collector. |
||
| Step 4 |
destination sup-eth0 Example: |
Specifies the collector destination. |
||
| Step 5 |
exit Example: |
Exit the command interpreter. |
||
| Step 6 |
buffer-drop monitor name Example: |
Specify buffer-drop monitor. |
||
| Step 7 |
collector collector-name Example: |
Add a collector. |
||
| Step 8 |
queue queue-value Example: |
Specify the queue to monitor buffer-drop. queue-value can be from 0 to 7. |
||
| Step 9 |
sampling rate rate or sampling microburst microburst-rate microburst-packets Example:Example: |
Specify sampling for buffer-drop. The value of rate can be from 10 to 16777215. A value of 1 for capture all. Specify sampling for microburst. The value of microburst-rate can be from 122 to 100000. The actual value will be the nearest lower value multiple of 122. The value of microburst-packets can be from 64 to 1024. |
||
| Step 10 |
filter ip access-list access-list-name Example: |
Configure the IPv4 filter. |
||
| Step 11 |
filter ipv6 access-list access-list-name Example: |
Configure the IPv6 filter. |
||
| Step 12 |
exit Example: |
Exit the command interpreter. |
||
| Step 13 |
buffer-drop system monitor monitor-name Example: |
Apply buffer-drop monitor. |
About Buffer Latency Capture
Buffer Latency Capture (BLC) allows you to capture information about packets experiencing high delays in the device. The original packets experiencing high delay are unaffected and forwarded normally. However, by using BLC, the device is able to capture the delay information for any packet experiencing a delay greater than a user specified delay threshold. The first 150 bytes of the original packet are sent along with metadata to the collector as an BLC packet. The device provides sampling mechanisms such as probabilistic capture and micro-burst capture that can be applied to the BLC packets.
BLC allows users to capture information about packets experiencing high delays in the device.
 Note |
The original packets experiencing high delay are unaffected and switched normally. Information about those packets may be sent to a collector for additional analysis. |
Information about the packet experiencing high delay can be sent to a remote collector or local CPU.
User has the ability to specify what the delay threshold should be that defines ‘high delay’.
For any packet experiencing delay greater than the specified delay threshold, metadata (shim header) is added to the packet experiencing delay and the first 150 bytes of the packet along with the metadata is sent to the collector
The metadata has more details about the packet that experienced delay, the format of metadata is listed below.
The device provides sampling mechanisms such as probabilistic capture and microburst capture which can be applied to the BLC packets.
BLC can be enabled/disabled per egress queue.
Buffer Latency Filter Types
|
Filter Types |
Filter_types gRPC Message Number |
|---|---|
|
FILTER_TYPE_BUFFER_LATENCY |
9 |
Configuring Buffer Latency Capture
Buffer Latency Capture allows the ability to capture packets exceeding a specified latency threshold.
Procedure
| Command or Action | Purpose | |||
|---|---|---|---|---|
| Step 1 |
configure terminal Example: |
Enter global configuration mode. |
||
| Step 2 |
hardware-telemetry buffer-latency Example: |
Enable buffer latency configuration. |
||
| Step 3 |
buffer-latency collector collector-name Example: |
Define a collector. |
||
| Step 4 |
destination sup-eth0 Example: |
Specify the collector destination. |
||
| Step 5 |
exit Example: |
Exit configuration mode. |
||
| Step 6 |
buffer-latency threshold thresh-name thresh-value Example: |
Define a threshold value. The thresh-value can be from 100 to 16000000 ns. |
||
| Step 7 |
buffer-latency record name Example: |
Define a buffer latency record. |
||
| Step 8 |
queue value interface ethernet slot/chassis latency-threshold t1 Example: |
Specify queue or queues to monitor buffer latency. |
||
| Step 9 |
exit Example: |
Exit configuration mode. |
||
| Step 10 |
buffer-latency monitor monitor-name Example: |
Specify buffer-latency monitor. |
||
| Step 11 |
collector name Example: |
Add a collector. |
||
| Step 12 |
record name Example: |
Add a record. |
||
| Step 13 |
sampling rate rate or sampling microburst microburst-rate microburst-packets Example:Example: |
Specify sampling for buffer-latency. The value of rate can be from 10 to 16777215. Specify sampling for microburst. The value of microburst-rate can be from 122 to 100000. The actual value will be the nearest lower value multiple of 122. The value of microburst-packets can be from 64 to 1024. |
||
| Step 14 |
filter ip access-list access-list-name Example: |
Configure the IPv4 filter. |
||
| Step 15 |
filter ipv6 access-list access-list-name Example: |
Configure the IPv6 filter. |
||
| Step 16 |
exit Example: |
Exit configuration mode. |
||
| Step 17 |
buffer-latency system monitor value Example: |
Configure buffer latency monitor. |
TCAM Carving for Flow of Interest
switch# configure terminal
Enter configuration commands, one per line. End with CNTL/Z.
switch(config)# hardware access-list tcam region hw-telemetry 128
Warning: Please save config and reload the system for the configuration to take effect
switch(config)# copy r s
[########################################] 100%
Copy complete, now saving to disk (please wait)...
Copy complete.
switch(config)# reload
This command will reboot the system. (y/n)? [n] y
switch# show hardware access-list tcam region | in hw-telemetry
HW Telemetry [hw-telemetry] size = 128
Feedback