ISPPlanet Cache Review Series - CacheFlow 545

Advanced Monitoring

Several methods are available to monitor CacheFlow events: an event log (stored locally or sent to a syslog server), email notification, and SNMP-based management. Management Events panels configure logging level (i.e., severe, resource, info, verbose), maximum log size and action, and the log host used when syslog forwarding is enabled. The event log can be examined locally with the Statistics Event Viewer panel (right).

This applet is ok for viewing log tails, but thorough analysis really requires the raw log, accessible from an unpublished URL.

The Events Mail panel (left)configures email addresses and an SMTP server used to forward daily performance summaries and notify about significant events (e.g., reload, NTP server reachability failure.) By default, notifications are sent to CacheFlow support for proactive problem resolution and MRTG performance monitoring for customers enrolled in the Cache Watch Monitoring program.

 

Management SNMP panels configure the agent's read, write, and trap community strings, MIB-II system objects, and up to three trap destinations. The CacheFlow can be monitored by any SNMP v1, v2, or v3 NMS; we monitored standard authentication and warmStart traps using CastleRock's SNMPc (right).

No cache-specific traps are generated by the current release, but enterprise traps, the RFC 2594 MIB, and the NLANR Proxy MIB will be supported in the 3.0 release. The current release has laid a solid foundation for integrated management; we expect the next release to build substantially on this foundation.

Introduction/The Platform	Predictive Caching
Four-Step Setup	Controlling Cache Services
Designing Your Cache Network	Viewing Cache Performance <
Managing the CacheFlow	Advanced Monitoring
Balancing Simplicity vs. Volume	> Product Support/Final Thoughts