Phone:400-882-3320
Guance 
Customer context
Peet’s Coffee is a coffee brand serving customers in China through stores and digital channels. Its published story describes a technology estate shared by store operations, online services, and enterprise users.
Digital workflows involved both store and online channels and depended on external services, so an incident could cross network, application, and third-party boundaries.
The story describes an IDC and AWS architecture linked through SD-WAN to stores, offices, and mobile workers, creating a need for a shared request-path and log view.
Hybrid-cloud and store paths crossed environments
Access across the IDC, AWS, stores, offices, and mobile endpoints required one view of network, application, and business dependencies.
Log storage and analysis were fragmented
Stores and online services generated continuing log data whose search, retention, and manual analysis were difficult to coordinate across systems.
Third-party service errors were hard to isolate
An external error required application performance, request-path, and error-detail evidence to establish the affected boundary.
Implementation
Draw application topology through a data gateway
The gateway ingested application traces and built a call topology where teams could review service relationships, response, errors, and environment context.
Continue from APM into third-party error details
APM related abnormal requests to external calls and error details, helping teams distinguish third-party, interface, and internal application causes.
Search store and online logs together
Logs entered a common analysis path and could be correlated with application traces and infrastructure state by service and time range.
What changed
Hybrid-cloud and store access paths became clearer
IDC, AWS, store, office-network, and application-call relationships could be reviewed in one investigation.
Third-party dependencies gained request-level evidence
Responders could continue from an abnormal service into its external call and error details.
Log analysis joined application investigation
Store, POS, and online-service logs no longer had to be searched without request and resource context.
Development and operations shared an event timeline
Topology, APM, logs, and infrastructure evidence provided a common basis for cross-team analysis.
