2. Agend
a
• Introduction
• What are the
core components
• Legacy and new
Architecture
• Deployment Flow
and Strategies
• Transitions
• Telemetry Ingest
• Conclusion
2
4. Secure
Network
Analytics
Behavioral modeling
Behavioral analysis of every activity
within the network to pinpoint
anomalies
Data collection
Rich telemetry from the existing
network infrastructure including
enhanced telemetry for
encrypted
traffic
analytics
Cisco XDR
Extended Detection and Response with Cisco
XDR. Advanced analytics extends local
detections with global intelligence and
integrations for accelerated response
Multilayered machine learning
Combination of supervised and unsupervised
techniques to convict advanced threats with high
fidelity
Secure Network Analytics
Endpoint Telemetry
Device and process insight with
flow telemetry from Cisco Secure
Client
4
5. Contextual
network-wide visibility
Agentless, using existing
network and cloud
infrastructure, even in
encrypted traffic
Predictive
threat analytics
Combination of behavioral
modeling, machine learning
and global threat intelligence
Automated
detection and
response
High-fidelity alerts prioritized
by threat severity with ability
to conduct forensic analysis
Cisco Secure Network Analytics
5
6. Multi-telemetry ingest and visibility
VPC, NSG flow logs
Secure
Network
Analytics
On-prem
Network
Telemetry
Admin Data
center
Networ
k
Users
On-premises
network
Remote Workers
Campus/
Branch
Public
Clouds
Cisco Firewall Log Data
Endpoint
Data (NVM)
6
7. Extensible Telemetry
Ingest
AnyConnect
Secure
Mobility Client
Identity
Services
Engine
AHGA/
ADC*
Proxy
Integration
*
Secure
Web
Appliance
Other
Web Proxies
ETA Capable Devices
Secure
Firewall
Flow
Sensor
NetFlow Enabled
Devices
IPAM DB Threa
t
Intel
Network
Telemetr
y
HTTP(S) Requests
HTTP(S) Responses
HTTP(S) URL
Custom HTTP(S)
Headers
Username
TLS Version
Key Exchange
Authenticatio
n Alg. MAC
Username
MAC Address
TrustSec Groups
OS Type
Process name
Process hash
Process account
Parent process
name Parent
process hash OS
Version Connected
interface
….
Flow Action
Translated
Port/IP SYSLOG
Connections
Malware events
File events
Hardware
events
L7 Application
HTTP Requests
HTTP Responses
SRT/RTT
TCP Flags
Payload
SRC/DST IP Address
SRC/DST Port
Bytes/Pkts Sent
Bytes/Pkts Received
…
(NetFlow, IPFIX)
Host
Group
s
VPC & NSG
flow log
transformatio
n via CTB
7
10. Secure Network Analytics Component Icons
flow data
(ETA Fields)
global
threat
alerts
Flow
Sensor
VM
VM
Non-NetFlow
enabled
equipment
ETA enhanced NetFlow
Hypervisor
with Flow
Sensor VE
proxy
data
NetFlow enabled
routers,
switches, firewalls
NetFlow
UDP Director
Additional Traffic
Analysis Software
(SIEM: Splunk)
Flow Collector
Manage
r
Threat
Intelligenc
e License
global
threat
alerts
telemetry for
encrypted
traffic analytics
(ETA)
End of Life
cloud-based
machine
learning
10
11. Secure Network Analytics components
Manage
r
Flow Sensor
Flow Collector
SMC VE (Virtual Edition)
SMC 2210
• SMC for Management and
Configuration supports:
• Up to 25 Flow Collectors
• 10000 Network Access
User sessions
• 15 concurrent managing
users
• Scale up to 6 Million
FPS in one
deployment
Flow Collector VE
FC 4210/FC5210
• Flow Collector is the center of
Data Collection and Analytics.
• Up to 25 FC per deployment
• Up to 240 000 FPS per FC
• Up to 6TB of Flow Storage
• Up to 1Million Host
Classified
• Up to 4000 Data Source per
FC
Flow Sensor VE
FS1210/FS 3210/FS4210
• Ingest SPAN to generate
telemetry
and contextual data.
• Up to 80Gbps per FS, Copper
and Fiber supported
interface,
• 1Gb, 10Gb and 40 Gb
monitor interfaces
11
12. UDP
director
UDP Director VE (Virtual
edition) UDPD 2210
Replicates UDP traffic and generates NetFlow
from SPAN traffic supporting:
• 1Gbps/10Gbps interfaces
• Up to 150,000 pps
Allows NetFlow, SYSLOG and SNMP data to be sent
transparently to multiple collection points
Provides additional flexibility and ease of
deployment
Secure
Network
Analytics
manager
NetFlow
Telemetry for
Encrypted
traffic analytics
(ETA)
NetFlow enabled
routers, switches,
firewalls
UDP
Director
Flow Collector
NOT FOR
SALE
12
13. Required core components
Flow rate
license
Secure Network Analytics manager
• A physical or virtual appliance that aggregates, organizes, and presents analysis
from flow collectors
• Central management for all Secure Network Analytics devices
• User interface to Secure Network Analytics
• Maximum 2 per deployment
Flow collector (FC)
• A physical or virtual appliance that aggregates, normalizes and analyze
telemetry and application data collected from exporters such as
routers, switches, and firewalls
• High performance NetFlow/SFlow/IPFIX collector
• Maximum 25 per deployment
Flow rate license
• Collection, management, and analysis of telemetry by Secure Network Analytics
• The flow rate license is simply determined by the number/type of switches, routers, firewalls and probes
present
on the network
• FPS estimation Tool
https://apps.cisco.com/cfgon/public/app/lancope/fpsestimator.jsp#/add-
items
Flow Collector
Secure
Network
Analytics
manager
13
15. flow data
(ETA Fields)
global
threat
alerts
Flow
Sensor
Hypervisor
with Flow
Sensor VE
VM
VM
Non-NetFlow
enabled
equipment
ETA enhanced NetFlow
proxy
data
NetFlow enabled
routers, switches,
firewalls
NetFlow
Telemetry
Broker
Additional
Telemetry
Destinations (SIEM:
Splunk)
Flow Collector
Manage
r
Threat
Intelligenc
e License
global
threat
alerts
telemetry for
encrypted
traffic analytics
(ETA)
Data Store
VPC
Secure Network Analytics Data Store Component &
CTB
End of Life
cloud-based
machine
learning
15
16. Secure Network Analytics manager
• A physical or virtual appliance that aggregates, organizes, and presents analysis
from
flow collectors
• Central management for all Secure Network Analytics
devices
• User interface to Secure Network Analytics
• Maximum 2 per deployment
Flow collector (FC)
• A physical or virtual appliance that aggregates, normalizes and analyze
telemetry and application data collected from exporters such as
routers, switches, and firewalls
• High performance NetFlow/SFlow/IPFIX collector
• Maximum 25 per deployment
Data Store (DS)
• A physical or virtual appliance that store data in a scalable, resilient
way.
• Maximum 12 per deployment (36 nodes)
Flow rate license
• Collection, management, and analysis of telemetry by Secure Network Analytics
• The flow rate license is simply determined by the number/type of switches, routers, firewalls and probes present on the
network
• FPS estimation Tool: https://apps.cisco.com/cfgon/public/app/lancope/fpsestimator.jsp#/add-items
Data Store Required core
components
Flow rate
license
Flow
Collector
Secure Network
Analytics Deployment
Data
Store
16
Manage
r
21. Deployment
Requirements
• IP addresses for appliances to be
deployed
• DNS Server IP(s)
• NTP Server IP(s)
• SMTP relay (if needed)
• Internal IP ranges in use/to
be monitored
Only for Data Store per node
• Non-routable IP Address from the
169.254.42.0/24
21
Device Information Communication Ports –NOT Full
LIST From (Client) To (Server) Port Protocol
Admin User PC All appliances TCP/443 HTTPS
Admin User PC All appliances TCP/22 SSH
All appliances
Network time
source UDP/123 NTP
Flow Collector SMC TCP/443 HTTPS
Manager Flow Collector TCP/443 HTTPS
Manager Flow Sensor TCP/443 HTTPS
Manager Internet TCP/443 HTTPS
Manager DNS UDP/53 DNS
Flow Sensor SMC TCP/443 HTTPS
Flow Sensor Flow Collector UDP/2055 NetFlow
NetFlow Exporters Flow Collector UDP/2055* NetFlow
NOT FULL List of
Ports
22. Deployment
Steps
First Time Setup
Appliance Setup Tool
• Interface SFP or BaseT Selection
• IP address Subnet
Configuration
• For Data Node 2nd IP non-
routable
• For FC Telemetry Selection
and UDP Port Definition
• Password Change
• IP address Subnet
Configuration
Verification
• SNA Domain and Type Type
(DS
or Not)
• DNS and NTP
Console
Http://IPaddres
s
Reboot is common between
steps
22
Removed 7.5
Less Restart
23. Cenralized Management
Connect With the
Manager
• Connecting to the Manager
• Will also Use the AST (Appliance
Setup Tool) (From FST in 7.5)
• After the AST Reboot
• Devices Connected
• Data Store Not Initialized
Initialize the
Data Store
• Go back to the Central
Manager console
• Initialize the Data Store
23
24. Smart Licensing Deployment Options
• Cisco product sends usage
information directly over the
internet. No additional
components are needed.
• Cisco products send usage
information to a locally
installed appliance.
• Periodically, exchange
information with Cisco to ensure
license usage is accurate.
• This synchronization can
occur automatically in
connected environments or
manually in disconnected
environments.
• Use copy/paste information
between product and
Cisco.com to manually check in
and out licenses.
• Functionally equivalent to
older node locking, but with
Smart License tracking.
Direc
t
On-
Prem
Offline
(not
recommended)
24
25. Licensing Notes
25
• After 90 Evaluation period ends the system will stop processing new flows
• Still functional with historical data, but new flow data will not be processed
• This is the ONLY hard enforcement used in Smart Licensing
• After a system is registered and the associated licensing periods expire or
are exceeded there is no hard enforcement
• The system will display banners informing users they are out of compliance,
but the system will still process flow
26. Flow estimation
• It is an estimated
Value unless you do a
PoV
• FPS license is based on
95th percentile, for 95%
of the time the FPS
actual is AT or BELOW
the stated amount
For every 1000 fps per day you need 1 GB storage at the Flow
Collector
26
28. What is the Data Store
• The Data Store is a new and improved
database architecture design for SNA
• Each individual Data Store appliance will
include a 3-Node database cluster
• Flow ingest by Flow Collectors is separated
from
data storage
• This distributed design enables scalable and
resilient data storage, providing retention
times of over a year
• Queries are handled by the Data Store,
effectively increasing performance across
all metrics by a significant magnitude 1 or more Flow
Collectors
Management
Console
3 or more Data
Nodes
28
29. With and Without the Data Store
1M FPS/90 days storage
• 16 total nodes: 8 data nodes + 8 Flow Collectors
(FC)
• Coupled Data collection & storage
• 10 total nodes: 6 data nodes + 4 FC
• Independent data collection &
storage
• Efficient and optimized data storage
1M FPS/90 days storage
Flow
Collectors
(FC4210)
FC5210
Data Base
Current Customer
Deployment
New Data Store
Deployment
Data
Nodes
FC5210
Engine
29
30. Data Resiliency
In addition to extending retention time, the Data
Store also introduces enterprise-class data
resiliency
1 or more Flow
Collectors
Management
Console
• Telemetry data is stored redundantly across nodes to allow
for seamless availability during single node failures
• Seamless availability for a Data Store deployments
1 2 3 4 5 6
3
o
r
m
o
r
e
D 30
31. Data Store Performance
Top Reports Non-DS Data
Store
Applications 5hr 47min 21min
Hosts 29hr 36min 19min
31
Ports 9hr 53min 19min
Protocols 6hr 50min 28min
Services 6hr 2min 20min
• Large Enterprise traffic, ran for 3 days
at 150,000 fps into two hardware
testbeds:
• FC5210 (Non-Data Store)
• 3-Node Data Store with a FC4210
• After 3 days, 19.4 Billion flows were
written to each testbed bed
32. Data Store Deployment
Single Switch
Architecture
32
Two Switch
Architecture
• eth2 or eth3 can be used for
internode communications
• Must be 168.254.42.x/24
• Provides resiliency for switch failure using
port channels and interconnected trunk
ports
• Uses both eth2 and eth3 for port channel
33. Data Store
Evolution
7.3.0
Data Store on
HW Data
Nodes is
introduced
7.3.1
Virtual Data Nodes
were added,
enabling virtual
deployment
7.3.2
Added new
telemetry
types, Firewall logs and
Remote worker
visibility, all NVM fields
7.4.0
Virtual Manager and
Flow Collector(s)
and
a physical Data
Store Support
added for ASA
firewall logs
7.4.1
Expand to Data
Store
• Single Node
• Multi-
Telemetry
• New Analytics
7.4.2
Transition to Data
Store
• Existing customers
can transition to Data
Store
Geo-Redundancy
• New peer site design
M6 HW Support
• SFP Interfaces
33
34. The single node Data Store
• Single node Data Stores can be either virtual
or physical appliances
• Supports up to 4 Flow
Collectors
• Easily expands to a full 3 node cluster, which
now supports N+1 horizontal scaling
• Note: A Data Store must consist of homogenous data nodes,
either all virtual or all physical appliances
Data Store
Single Data
Node
Virtual or physical
Expand/Scale as
needed with
FCs
Single node virtual Data Store scales to 225K FPS
Single node physical Data Store scales to 500K
FPS
34
37. Redundancy – High Level Design – Non-Data Store
VIP
NetFlow/IPFIX (UDP 2055)
Heartbeat
Signal
Sent over
network
HTTPS
SMC
Primar
y
(active)
Broke
r
Primary
(active)
FC
Primary
(active)
SPAN
SMC
Secondary
(active)
Broker
Secondary
(passive)
FC
Secondary
(active)
FS
Secondary
(active)
FS
Primar
y
(active)
Exporter
s
Actual
Redundancy
37
38. Redundancy – Notes– Non Data Store
38
• SMC redundancy follows active – active (but no change)
• Flow Collector redundancy is active active and done by
design
• Flow Collector redundancy required double licensing
• CTB help in achieving the flow collection redundancy
• Flow Sensor redundancy is active - active
39. Resilient central storage for multi-geo ingestion
• Flow Collector consolidates redundant
flow
date into context rich bi-flow records
• Highly efficient compression minimizes
WAN impact when backhauling
telemetry data
• Telemetry data is stored redundantly
across data nodes to help ensure data
availability even during a data
node failure
• Redundant inter-connection switches, help
to ensure the Data Store stays in
operation during network
upgrades and unplanned outages
39
40. Redundancy – New Architecture
Requirement: Geographical
redundancy while minimizing footprint
Solution: Peer Sites
• Primary deployment is associated to a peer site
where configurations are sync’d
• Both sites run and operate independently,
allowing
great flexibility to meet customer operational
reqs
• Site telemetry is sent to both primary and peer
sites.
• Primary site can be robust HW appliances where
peer site is smaller virtual deployment
reducing OPEX
• Peer sites based on Active/Standby
Managers design, and is supported within
peer sights for large Enterprises
demanding full redundancy 40
41. What are the gotchas?
41
• Java/Swing client is not supported with Data Store
• BU is actively working to close reporting and data visibility
gaps
• Peer Site sync is manual
• 3+ DC designs are not supported today
• BU: Investigating extending peer site for this purpose
• Multiple Data Stores are not supported by a single Manager
• Converged Analytics cannot support multiple
domains, it runs on one domain at a time
43. Storage
(Archival, Audit)
3rd Party Cloud Services
(SIEM/Datadog/ServiceNow)
3rd Party On-Prem
(Home grown Data Lakes, Live Action, SevOne)
Cisco Cloud Services
(Secure Cloud Analytics, CDO, SecureX, SSE*)
Cisco On-Prem
Platforms
(DNA Center, Secure Workload)
CTB
Distributed
Nodes
Broker
Transform
Filter
Anonymize
API
Telemetry Sources Telemetry Destinations
Cisco Secure Network
Analytics
NetFlow /
sFlow
Syslog
VPC Flow Logs
SNMP Trap
Application
Sources
…and more
Sources = Network,
Application, or Cloud provider
points of telemetry egress.
Endpoint
(NVM)
43
*integrations under
investigation
44. Cisco Telemetry Broker Democratizes Telemetry Data
Quickly enable PoV/onboarding of non-incumbent
tools
Replicate
Telemetry
Control Costs: Only index high value
data
Brokering
Compliance: Keep low value data in low-cost
storage
Increased visibility of legacy
sources
Legacy protocol to Modern
tool
Transforming
The ability to transform data
protocols from the source to the
destinations protocol of choice
Modern protocol to Legacy
tool
Increased visibility into modern
sources
Route
Telemetry
Let teams run the tools of their choice without
deploying new agents/collectors
Filter to
Drop
and
Segment
The ability to route and replicate
telemetry data from multiple
source locations to multiple
destinations
Filtering
The ability to filter data being
replicated to enable fine grain control
over what destinations ingest and
analyze
44
45. Components of the Telemetry Broker
• CTB Manager node:
• Only one manager is deployed and can
manage
multiple Broker nodes*
• Maintains the policy/rules for the broker
nodes enabling central management from
one view
• If the manager goes down, broker nodes
continue to process telemetry
• Backup configurations are created for recovery
• CTB Broker node:
• Where the telemetry brokering work occurs
• Can be deployed closest to telemetry
sources
Manager
Broker
Node Broker
Node
Broker
Node
Management Network Monitoring Network
Broker
Node
*A single Manager supports up to 10 Broker
nodes
45
46. Minimum requirements for
a Cisco Telemetry Broker
Node:
• CPU: 2 cores (1 Gbps) or 5 cores
(10Gbps)
• Memory: 4 GB (1 Gbps) or 8 GB (10
Gbps)
• Storage: 20GB
Minimum requirements for a
Cisco Telemetry Broker
Manager:
• CPU: 2 cores
• Memory: 8 GB
• Storage: 50GB
Cisco Telemetry Broker
Can also be deployed on a UCS
server!
Versions
6.7 or 6.5
46
**See notes for more
details
https://cs.co/
telemetrybroker
47. Hardware broker node
• Supports 300k FPS capable of uploading to Cisco XDR
• Dedicated 10GB Management and Monitoring interfaces
• 16 x 16 GB DDR4 3200 memory
• 6 x 600GB 10K RPM RAID6(data), 2 x 240GB Data M.2 RAID1 (OS) storage
• 2 x Processor AMD EPYC 7313 16C/32T @ 3.0Ghz or boost 3.7Ghz processor
47
48. High Availability
• HA Configurations are supported
for
Cisco Telemetry
Broker
• Simply scale more brokering nodes
to provide for resiliency
• HA Broker nodes will operate in
standby mode until their associated
active node goes down
• Broker nodes can be geo-
distributed with the manager
centralized
• Broker nodes operating in standby
mode will not process any
telemetry and will not
incur any additional
licensing cost
M
B B
B
B B
B
48
50. Migrating from a
UDPD
• Cisco Telemetry Broker improves
upon the successful feature set of
the UDP Director
- CTB improves performance,
simplicity, and offers new feature
functionality
• Cisco Telemetry Broker can use an
existing configuration file from
UDPD to seamlessly integrate
existing forwarding rules
• Device architectures are
different
- Account for the addition of
Brokering Nodes in an existing
design
- Account for new licensing model
UDP
Directo
r
Brokerin
g
Filtering Transformin
g
Brokerin
g only
Secure
Network
Analytics
SIEM Cisco
DNA Center
Secure
Network
Analytics
SIEM Cisco
DNA
Cent
er
Secure
Network
Analytics
SIEM Secure
Network
Analytics
Logs
IPFIX
VPC
Flow logs
Cheap Storage
IPFIX
IPFIX
50
51. Data Store Transition
No need for forklift upgrades to achieve
success!
Upgrad
e
Softwar
e
2
Hardware
generations
supported 4K and 5K
Re-use
Managers
Flow
Collectors
Flow Sensors
51
No other vendor in
the market supports
this model
52. Transitioning to Data Store
Today
FC
• Data Store is added to the existing deployment
• Upgrade existing FC (engine) to send new
telemetry to the Data Store
• The FC (DB) will stay in existing format
• Manager communicates with the Data Store to run
reports and flow searches for recently ingested
telemetry
• Manager queries FC (DB) for older reports and searches
FC
• After FC (DB) retention time has expired,
DB portion is decommissioned
• SMC no longer queries FC (DB)
• For virt FC, and FC42xx/FC43xx FC (DB)
resources are returned to system to
optimize FC performance (up to
three times faster)
Not the 5K DB Node
End State
• FC 4210 DDS
• FC 5210 DDS
• FC 4200 DDS
• FC 5200 DDS
FC
Transition
State
52
53. Transition Steps
53
• From the Manager web UI
- Step 1: Create a Data Store domain
- Step 2: Setup sync between non-Data Store domain to Data Store
domain
- Step 3: Sync the domains
• From the Manager CLI (SystemConfig as root)
- Step 4 – Add the data node(s) to Central Manager
- Step 5 – Enable SSH on the Data Store
- Step 6 – Initialize the Data Store
- Step 7 – Pick the flow collector and domain for
transitioning
- Step 8 – Acknowledge the flow collector transition
• From the Manager web UI
- Central Manager>Inventory tab will show a transition flag (Data Store Transition) next to the flow
collector
- Central Manager>Data Store tab will show “Oldest Record (days ago)” for NetFlow, NVM and Firewall
logs.
- Once there is 30 days for each then the transition can be completed
• From the Manager CLI (SystemConfig as root)
- Step 9 – Select Data Store then Complete Transition and then the flow collector to transition
- Step 10 – Acknowledge to complete the transition (note all old data on the flow collector will be
deleted)
Transitio
n
Setup
Initiate
Transition
Monitor
Transitio
n
Complete
Transition
55. Multi-telemetry ingest and visibility
VPC, NSG flow logs
Secure
Network
Analytics
On-prem
Network
Telemetry
Admin Data
center
Networ
k
Users
On-premises
network
Remote Workers
Campus/
Branch
Public
Clouds
Cisco Firewall Log Data
Endpoint
Data (NVM)
55
56. Netflow Required Fields
56
The fields that SNA requires to ingest flow
are: Field NetFlow Element ID Configuration Example Required Field?
NF_F_PROTOCOL 4 match ipv4 protocol Yes, Key Field
NF_F_SRC_ADDR_IPV4 8 match ipv4 source address Yes, Key Field
NF_F_DST_ADDR_IPV4 12 match ipv4 destination address Yes, Key Field
NF_F_L4_SRC_PORT 7 match transport source-port Yes, Key Field
NF_F_L4_DST_PORT 11 match transport destination-port Yes, Key Field
INPUT_SNMP 10 match interface input Yes, Key Field
SRC_TOS 5 match ipv4 tos Yes, Key Field
OUTPUT_SNMP 14 collect interface output Yes, Key Field
NF_F_IN_BYTES 1 collect counter bytes Yes, Key Field
NF_F_IN_PKTS 2 collect counter packets Yes, Key Field
NF_F_LAST_SWITCHED 21 collect timestamp sys-uptime first Required; for calculating duration
NF_F_FIRST_SWITCHED 22 collect timestamp sys-uptime last Required; for calculating duration
NF_F_TCP_FLAGS
57. Netflow Required Fields
57
ETA Fields ETA Fields ETA Fields ETA Fields
44940 ipv4 idp This is Initial Data Packet; used for crypto audit
44941 ipv4 splt
SPLT - Sequence of Packet Lengths and Times ;
malware detection
44944 ETA Byte Distribution; malware detection
NBAR Data NBAR Data NBAR Data NBAR Data
12235 match application name NBAR application data
45003 match application name NBAR application data
Additional Fields Additional Fields Additional Fields Additional Fields
initiatorOctets 231 collect connection initiator This field is useful to determine the flow initiator
natEvent 230
Without this field we cannot get firewall events for the
flow (denied, accepted, etc)
58. VPC Flow Logs to IPFIX
10010
11011
Flow Logs Transformed to
IPFIX
Flow
Logs
Secure
Network
Analytics
On-Prem
Networks
NetFlow
IPFIX
100101
101011
• Cloud Flow Logs from AWS and Azure
provide insight into the activities of hosts
residing within cloud environments
• Meta data from Flow Logs centers around
the network activity, similar to
IPFIX/NetFlow
- There are 25 total fields provided in Flow Logs
- Fields provide insight to network metadata
as well as metadata associated with the
VPC/NSG
• CTB pulls Flow Logs from AWS S3 buckets and
Azure BLOB storage via secure HTTPS
connections and transforms the telemetry to
IPFIX
- Once the VPC flow is transformed it
is then forwarded to consumers
58
59. Complete and continuous remote worker
visibility The Cisco Secure Client (AnyConnect Secure Mobility Client) caches all
network traffic telemetry records, even when users are not using a VPN
• On-network flows (collected when VPN connected) – real time
• When user connects to VPN all stored NVM flow data is sent to the Flow
Collector
• Can be configured for burst or chunks and adjustable cache size
• Detections are carried out on the NVM flows
(Behavioral, Customer Security Events and Converged Analytics)
• Note a flow search does not show NVM specific fields
• Off-network flows (collected when VPN not connected) – cached late
arriving
• Can view the historical NVM flow data using the NVM endpoint traffic reports in
Report Builder
• No detections are applied to off-network traffic
nvm_to_flow_cache
nvm_filter_untrusted_flow 59
60. No Endpoint VM Since 7.3
Recap of all NVM telemetry records retained
User Endpoints with
AnyConnect
Secure Mobility
Client
Start Time*
End Time*
Source IP*
Source Port*
Destination IP*
Destination Port*
Bytes Sent*
Bytes Received*
Packet Count*
(derived)
Protocol*
Interface Info UID
Interface Index
Interface Type
Interface Name
Interface Details List
Interface Mac Addr.
UDID
User
User Account Type
Agent Version
Virtual Station Name
OS Name
OS Version
OS Edition
System Manufacturer
System Type
Process Account*
Process Account Type
Process ID
Process Name*
Process Hash*
Process Path
Process Args
Parent Process ID
Parent Process Account
Process Account
Parent Process Name*
Parent Process Hash*
Parent Process Path
Parent Process Args
Host Name
DNS Suffix
Module Name List
Module Hash List
Parent Process Name
Parent Process Hash
Flow
Collecto
r
Data
Store
NVM
Telemetry
• Session
• Interfac
e
• User
• OS
• Process
* NVM telemetry records available within non-Data Store
deployments 60
61. Store Cisco Firewall logs on premise with Data Store
Cross launch from FMC with context
into Secure Analytics and Logging
dashboard
Make data available to FMC via
APIs for supporting remote query
100,000k eps (8.65 Bn/ day)
support for
+30 days using full data store
architecture
61
62. FMC pivots directly to the Data Store with enhanced
context
• Contextual pivots from Firepower Management Center to the event
viewer
optimizes SecOps workflows by automatically filtering on events of
interest
Remote Query API do not support ASA
Events
62
63. Intelligent viewer provides access to all Firewall data
• Select custom
timeframes going back
across over any
retention time
• Filter exclusively on
Security Events and
use per column filters
to quickly isolate data
of interest
• Create custom view
to tailor content
based on columns
shown
• Use Summary to
identify trends and
outliers
• Export any view to CSV
for archiving or to
further forensic
investigations 63
64. Secure Network Analytics detection architecture
NetFlo
w
Devices
Secure
Firewall
Secure
Client
Flow
Collector
Data
Store
Converged Analytics
with MITRE ATT&CK
mappings
VPN and
Split
Tunnel
off-network NVM and
FW event logs are saved
to the Data Store without
analysis
Remote worker off-network
(NVM)
IDS Events
Malware Events
LINA Events
Connection
logs
Cisco
XDR
XDR
NetFlow
Connectio
n logs
Promote
d Events
Behaviora
l security
events
SRC/DST IP Address
SRC/DST Port
Protocol
Bytes/Packets Sent
B
…yte
.s/Packets
Received
FC Analytics
Engines Custom
Security
Events
Network
Context based
Detections Firewall events can be
sent to Cisco XDR from
FMC
in addition to being
stored
on-prem
Mana
ger
Firewall detections
based on Firewall event
context planned
Will Change
64
65. SNA Firewall Logs Detections
Secure
Firewall
Flow
Collector
Data
Store
IDS Events
Malware Events
LINA Events
Connection
logs
Connectio
n logs
Behaviora
l security
events
Custom
Security
Events
Ingest Convert Detect
Firewall Logs
Ingested on
port 8514
65
Connections
Logs
Converted To
Flow
Apply
Behavioral and
Custom
Detections
Flows converged from
Firewalls logs to
Netflow do not count
against the FPS
license. SAL is licenses
per GB/Day already
Firewall Logs from a device
should not be sent at the
same time with Netflow, that
will create wrong ByteCount.
Now ! SNA detections with
a firewall only as a
telemetry source.
(Connection End)
66. SNA Firewall Logs Detections
Leverage
Analytics to
trigger
Behavioral Alerts
Flow Denied Security Event in SNA triggered on traffic from Firewall Logs
Customized
Alerts with
Custom
Security
Events
66
67. SNA Firewall logs to Detections Configurations
67
Configuration
• sal_enable = 1
• sal_to_flow_cache = 0 (default) Put it to 1 to enable
conversion
• sal_port = 8514 (default) Ports should not be overlapped
with
other ports 2055 for Netflow and 2030 for NVM
Troubleshooting
• /lancope/var/sw/today/logs/sw.log
05:00:02 S-per-t:
05:05:00 S-per-t:
Current sal_event, Input: 0, Decoded: 0, Output: 0, Ignored:
0
Current sal_event, Input: 3325, Decoded: 3325, Output:
3325,
Ignored: 0, Dropped: 0, To_Flow: 1578 this period
Dropped: 0, To_Flow: 0 this period
05:10:00 S-per-t: Current sal_event, Input: 4411, Decoded: 4411, Output: 4411,
No Pivots are available to FMC Not Available for
DDS
68. NVM Detections
Secure
Client
Flow
Collector
Data
Store
FC Analytics Engines
VPN and
Split Tunnel
Behaviora
l security
events
Custom
Security
Events
No Conversion of NVM to
flows required, Flows
then goes through the
detection engine
All Detection from
behavioral analytics can be
applied including data
movement.
Alert on Process Names and
Hashes
with CSEs in addition to all other 68
One More way to Deploy SNA
without having to be restricted to
network flow to get detections
69. NVM Detections Configurations
69
New Install
nvm_enable = 1
nvm_to_flow_cache = 0
(default) nvm_port = 2030
(default)
NVM flows can be seen in flow search and Report
Builder when nvm_to_flow_cache is enabled
NVM flows can be seen in only Report Builder when
nvm_to_flow_cache
is not enabled
Troubleshooting
/lancope/var/sw/today/logs/sw.log
/lancope/var/logs/containers/svc-db-ingest.log
70. New Detections and Alerts in Converged Analytics
4 New Alerts from
Secure Cloud Analytics
• LDAP Connection Spike
• Outbound LDAP Spike
• Protocol Forgery
• Repeated Umbrella
Sinkhole
Communications
2 New Observations
from Secure Cloud
Analytics
• ISE Session Started
Observation
• Umbrella Sinkhole
Hit
Observation
70
71. Dynamically maps entities by role
Functional
modeling
Type based
modeling
Android
Citrix PVS server
Windows workstation
Mail server
Medical imaging client
Remote desktop server
DNS server
VolP client
Apple iOS
Web server
Wireless LAN controller
Domain controller
…over 50+ entity roles are supported !
• Automatic role classification available on a new report leveraging the new converged analytics capability
• Roles are available out of the box with no tuning and provide details about devices on the network for
investigation
Roles include:
71
72. Device Report Traffic enhanced with automatic filters
• Select any time on the traffic statistics
graph and see results dynamically filtered
in the flow table
• Accelerates investigation of traffic
anomalies
• Immediately correlates chart events with
actual flows attributing to the event
72
75. Where To Enabled
Telemetry ?
Flow
Sensor
Hypervisor
with Flow
Sensor VE
VM
VM
Non-NetFlow
enabled
equipment
ETA enhanced NetFlow
proxy
data
NetFlow enabled
routers, switches,
firewalls
NetFlow
Telemetry
Broker
Additional
Telemetry
Destinations (SIEM:
Splunk)
Flow Collector
Manage
r
telemetry for
encrypted
traffic analytics
(ETA)
Data Store
VPC
75
76. The more you enabled the more you see
Visibility into traffic going through
these devices to the internet or to
the main data center. Firewalls
could provide NAT information
Visibility into traffic going through
the core to the Internet or to the
DC and Campuses
More visibility into user traffic
from one VLAN to another or
even from port to port
Edge Devices at Campus
Edge Devices at
Branches
Core Switches
Access and
Distribution
76
77. Add A Flow Sensor
At the EDGE Get Application layer
visibility into your internet
traffic (URL and APPs)
Visibility into VM traffic and
additional network use
cases with RTT and SRT
Legacy Networks
visibility where flow is
not available
VM
VM At the
Hypervisor
Non Flow Capable
Networks
77
78. What Can A flow Sensor Do
Virtual or physical appliance that produces telemetry for
network
infrastructure incapable of generate NetFlow natively
Provides additional security context to enhance Secure
Network Analytics security analytics
Additional information gathered
• ETA enhanced NetFlow
• TLS Finger Printing
• Layer 7 application data
• URL information for web traffic
• TCP and ICMP flag details
• RTT (Round trip time)
• SRT (Server response time)
• Retransmissions
• X-Forwarded headers from web load
balancers
Non-NetFlow
enabled
equipment
Secure
Network
Analytics
manager
Flow Collector
VM
VM
Flow
Sensor
Hypervisor
with
Flow Sensor VE
ETA Enhanced NetFlow
78
79. Visbility at the Endpoint Level
User Endpoints with
AnyConnect
Secure Mobility
Client
Start Time*
End Time*
Source IP*
Source Port*
Destination IP*
Destination Port*
Bytes Sent*
Bytes Received*
Packet Count*
(derived)
Protocol*
Interface Info UID
Interface Index
Interface Type
Interface Name
Interface Details List
Interface Mac Addr.
UDID
User
User Account Type
Agent Version
Virtual Station Name
OS Name
OS Version
OS Edition
System Manufacturer
System Type
Process Account*
Process Account Type
Process ID
Process Name*
Process Hash*
Process Path
Process Args
Parent Process ID
Parent Process Account
Process Account
Parent Process Name*
Parent Process Hash*
Parent Process Path
Parent Process Args
Host Name
DNS Suffix
Module Name List
Module Hash List
Parent Process Name
Parent Process Hash
Flow Data
Store Collector
79
NVM
Telemetry
• Session
• Interfac
e
• User
• OS
• Process
* NVM telemetry records available within non-Data Store
deployments
80. Store and Analyse Firewall Logs
Min Supported
Version
Notes
FMC 6.7
Older versions are
supported but Cross-launch
will not be available
FTD 6.4
SMC 7.3.0 SMC VE or SMC 2210
SAL On Prem 1.0.0
An application needs to be
installed separately from
the SMC 7.3.0 install
Secure
Firewall
Flow
Collector
Data
Store
IDS Events
Malware
Events LINA
Events
Connectio
n logs
Behaviora
l security
events
80
Custom
Security
Events
81. Analyze your Cloud Data by adding CTB
srcaddr
dstaddr
srcport
dstport
protocol
packets
bytes
start
End
tcp-flags
sourceIPv4Address or sourceIPv6Address
destinationIPv4Address or destinationIPv6Address
sourceTransportPort
destinationTransportPort
protocolIdentifier
packetDeltaCount Secure
octetDeltaCount Netwo
flowStartSeconds
flowEndSeconds
tcpControlBits
Cisco Telemetry
Broker
Transformation
VPC Flow Logs IPFIX
rk
Analytics
81
82. Summarize The Telemetry Design Use Cases
Visibility into traffic
going through the
Network east West
extend it by
enabled at other
layers
82
Analyze and
Store your
firewall logs
and NAT
information
Get to the
endpoint
process user
and interface
level
Get visibility
into your cloud
environment
by leveraging
CTB
84. Secure Network Analytics integrations
Proxy
Web APP, web URL and user info
External lookup
Extended analytics, threat
investigation
DNA Center
Automated setup and
deployment
XDR
Threat hunting and
response
Secure Client
Process and endpoint
visibility
PAN
Application and user
identity
Identity Services Engine
User identity, device
identity, mitigation and
response
API
Automated and
customized configuration
and reporting
Secure
Network
Analytics
84
85. Secure Network Analytics and network access integration
Network Access services
and classification
Secure
Network
Analytics
Secure Network Analytics
visibility
Device Id
Domain
Id Active
Start active
time Endpoint
IP Username
SGT Tag
Trustsec name
Last update
time
InterfaceDevice
PortId
InterfaceDeviceI
p
Vlan
MAC address
Session ID
Active
Start active
time
Username
Last update
time
Secure Network Analytics also integrates with
ISE-PIC using pxGrid to get endpoint
contextual information
Secure Network Analytics integrates with ISE
to get mitigation capabilities and apply
different ANC policies to an endpoint
Info from
ISE
Info from ISE –
PIC
Cisco Identity
Services Engine
pxGrid
85
87. Secure Network Analytics is a
comprehensive data source
Source Information
Method SMTP
To inbox & ticketing
systems
Data exporter
Usually to
SIEMs
SYSLOG
To log aggregation
SOAP web services API
REST API
Alarms
Notification
s
Normalized flow
data
Data export from FC
APIs
Configuratio
n
and
reporting
Flow Collector
Manager
87
88. Improving on-prem NDR with Cisco
XDR Cisco XDR
Enrichment Requests from
manual investigations or
auto- mated from event
correlation
Alarms and
Events sent
to XDR
analytics
Tiles to
Contro
l
Center
Optional: Send
flows to XDR
analytics via CTB or
FC
Secure Network
Analytics
Cross correlation of data
Correlation of NDR findings with
other detections mechanisms
including EDR based detections,
email and others
88
Impact Analysis
Understand the Impact of an
incident
leveraging XDR incident Manager
Reduce the time to
respond
Reducing the time to
response leveraging XDR
automation and the multi
responses capabilities
Extend response
capability
Expand NDR response
capabilities with multiple
technologies through XDR
integrations with Cisco and 3rd
party technologies
89. Data Enrichment From SNA to
XDR
Events details are sent
with relationship
indicators for some alerts
when available
89
Security Events
Contribute into XDR
investigations
Security
Events
Investigation
Configuration
Limit
s
92. Rule
What alarms?
Action
What to do with alarms?
Response Manager
Logic
If Condition is met, then trigger
Response
Automatic
Response
Automate responses by
defining rules and
applying actions
92
New Alerts
option in
Response
Manger.
Converged
Analytics
needs to
be enabled.
Webhooks
supported for Alerts
(Converged
Analytics) and
Alarms.
Not customizable
for either.
New Actions for Converged
Analytics Alerts.
93. Rules With Flexible Conditions
93
Define Rules with
Multiple Conditions
• Granular control with
complex rule
triggering
conditions
Use Multiple
Actions:
• Select 1 or more actions
to be executed once
rule is matched
and alert is open
• Select 1 or more actions
to trigger when the
alert is closed
95. Most Utilized
Resources
98
SNA Resources Secure Analytics Videos
http://cs.co/SecureAnalyticsVideo
s
Detection:
Secure Analytics Detections Demo
playlist
Design Guide:
SNA Data Store Design Guide
FPS Estimator:
FPS Estimator
Training Center:
Secure Network Analytics Training Center -
Use
Cases
97. FC4300 in a Non–Data Store Deployment
• The Flow Collector does everything: ingestion, storage, and query.
• Retention & performance depend on appliance sizing:
• FC4300 can ingest up to ~4M flows per second (depending on
model and license).
• Stores flow records locally (on its own disks).
• Query performance is fine for short- to mid-term history (days
to a few weeks).
• As flow volume and query size grow, performance slows because
the collector handles queries + ingestion simultaneously.
• Best for small to medium deployments where:
• Flow volume is moderate.
• Long-term retention isn’t required.
98. FC4300 in a Data Store Deployment
• Flow Collector handles ingestion only and forwards flow data
into the Data Store cluster.
• Data Store handles indexing, long-term retention, and
queries.
• Benefits:
• Higher retention (months/years).
• Faster queries on large datasets because SMC queries the
Data Store, not the Flow Collector.
• Flow Collector resources are dedicated to ingestion,
improving stability at high flow rates.
• Supports scalability → multiple FC4300s feeding the same
Data Store cluster.
• Best for large enterprises, service providers, or compliance-
driven industries needing forensic data for long periods.