Network Performance Monitoring
Horizon Networks SuperNM Series
Horizon Networks SuperNM Series of network intelligent analysis products adopt the latest multi-core multi-processor technology. The visibility of the end-to-end application delivery correlation of SuperNM products is combined with network behavior analysis, which can analyze the behavior of TCP layer on each hop path of application or the behavior of all TCP based applications on a core node, and in case of failure, it can find and locate problems in time, so as to provide means for it operation and maintenance.
Super processing ability
Full traffic monitoring
N + B solution
“Super processing ability”
The original SPOS super parallel processing architecture is adopted. The maximum performance of a single device is 80Gbps, which reaches the industry-leading packet capture processing performance. The Ethernet packet is captured in real time, and the granularity of network performance analysis is achieved in milliseconds.
“Full traffic monitoring”
Combined with SuperAD products, it can realize the network performance monitoring of the real IP on the application server side of the data center, without additional SSL traffic decryption.
“N + B solution”
The only NPM + BPM unified architecture solution in the industry can realize real-time performance analysis from L2 to L7, and support business decoding, transaction Association and transaction tracking analysis. When analyzing problems, drill down from business data layer by layer to the original data package.
The alarm information generated by machine learning based on historical data can minimize false positive and false negative, effectively prevent the overall performance level of the business from decreasing, and help the operation and maintenance team more easily meet the operation and maintenance requirements of the core business network.
Fully support VMware, KVM, Openstack virtualization environment and Docker environment.
Intelligent report provides basis for business decision
The original packet level backtracking analysis can quickly locate the location of network failure, performance and abnormal traffic
Automatic learning modeling, intelligent early warning system to detect performance risks in advance
Realize large screen display of operation and maintenance, KPI scoring mechanism, and intuitive understanding of network and business health status
- Data processing
- Traffic analysis
- Application analysis
- Data Association
- Message extraction
IP subnet: segment grouping (IP Site, Site) according to a single IP address and IP subnet. Network traffic can be monitored and analyzed according to IP site. Public IP address is aggregated by province
MPLS: split the MPLS header of the data message and store VxLAN: split the VxLAN header of the data message and store the port convergence again: analyze the traffic aggregation of any multiple collection ports, and distinguish the uplink and downlink
System score: define the weight of each KPI of the application system, display the system health degree in the form of full score of 100 points, and reflect the KPI and value of score reduction through drilling.
Service topology: define the service topology, dynamically display the response time, network delay, retransmission, concurrency, traffic, alarm and other KPIs of business system flow in key equipment, link and application nodes.
For the device, such as the device topology, the difference value of the device's retransmission delay, dynamic packet loss, and so on.
Drill through operation: drill through the KPI data of any 1 minute to drill down the detailed KPI of TCP session within 1 minute.
Multi dimensional analysis: traffic query is conducted by sorting "application / Protocol", "TCP port", "IP address pair" and "IP address". It can be selected any IP address pair to retrieve the original network data message.The IP address pair of any query record can be selected to plot the trend chart of all KPIs related to the load, TCP performance parameters and TCP availability parameters of the IP address pair in the query time.
IP search: global search can be performed according to IP, and the traffic information of the IP as server and client on each interface can be statistically displayed. VLAN statistics: it supports traffic query statistics and sorting by VLAN as a unit, and can list the traffic ranking of hosts and communication pairs of a single VLAN, as well as the load, performance and availability curves of each VLAN. Site statistics: it supports traffic query statistics and sorting by site, and can list the traffic ranking of hosts and communication pairs of a single site. Quick application definition: in the traffic query results, it supports to quickly add servers and ports in unknown application traffic as custom applications. Host name management: support IP address custom naming, and can display IP address or host name in combination. Long connection analysis: it can automatically identify long connections, define the time limit of long connection identification, and provide the connection start time of TCP session in the TCP long connection list. Peak response analysis: it can list the sessions with the highest response time per minute for each rule, and record information including detailed parameters of all load, TCP response time and availability.
HTTP analysis: support real-time analysis of HTTP protocol application KPI: URL, request method, error code, application response time, total page transmission time. And it can extract single interactive network data message.
Oracle analysis: support real-time analysis of Oracle protocol application KPI: Oracle statement, error code, application response time, total statement response time. And it can extract single interactive network data message.
HTTPS decoding: support to decode the HTTPS traffic by importing SSL key.
Simultaneous data association: the peak value, valley value and baseline data of the previous cycle can be superimposed and displayed on the selected KPI sequence diagram.
Application network association: for HTTP and Oracle protocols, a single application interaction can be associated to drill the TCP session carrying the application interaction, and a single TCP session can drill each application interaction contained in it.
The alarm can be related to a single TCP alarm session.
Cycle baseline: taking each application rule as the baseline generation unit, the baseline cycle and granularity can be set, and the exclusion date and business time period can be defined
Security alarm: supports the baseline and threshold alarms of the following security events: TCP syn storm, TCP connection rejection, TCP port scanning, ARP broadcast storm, DDoS attack device alarm: supports the baseline and threshold alarms of KPIs such as device delay, packet loss, concurrency difference, retransmission difference, etc
Report export: support PDF export
Report sending: support mail sending