In a recent survey, research participants highlighted the importance of enhancing visibility into their data center network fabrics and WAN edge connectivity services. Many expressed the need for real-time data to effectively monitor AI networks and optimize network performance.
Observability tools often rely on SNMP polling to collect network metrics, but shorter polling intervals can impact both network and tool performance. Sixty-nine percent of respondents emphasized the necessity for real-time infrastructure monitoring to address issues like AI traffic bursts that may go unnoticed with traditional polling methods.
Additionally, 51% of participants indicated a need for more real-time network flow monitoring, especially in cloud environments where technologies like VPC flow logs may not offer sufficient data granularity. To bridge these visibility gaps, network teams may need to incorporate real-time packet monitoring into their observability strategies.
Furthermore, network teams are looking for observability tools that can intelligently identify AI applications in network traffic. By monitoring AI application performance, optimizing network configurations for AI traffic, and detecting any unauthorized AI adoption, these tools can help ensure smoother operations and enhanced network security in the age of artificial intelligence.
