Monitoring the Microsoft Edge Transport Servers

eG Enterprise offers a specialized Microsoft Exchange Edge Transport model that provides real-time insights into the performance of Edge Transport servers.

layermodel-final

Figure 1 : Layer model of the Microsoft Exchange Edge Transport server

Every layer of Figure 1 reports a wide variety of statistics that enable administrators to quickly find answers to the following critical performance queries:

  • Is the Active Directory cache adequately sized to handle requests from the Edge transport server?
  • search requests to any domain controller fail owing to a bad network link or the non-availability of the domain controller?
  • Were any LDAP fatal errors experienced while communicating with a domain controller?
  • Did too many bind calls to any domain controller fail?
  • Is any domain controller responding too slowly to read and search requests?
  • Is the server experiencing processing bottlenecks? Are there any lengthy message queues on the server? If so, which ones?
  • How effective is the Recipient filter agent? How many requests per second were rejected by the Recipient Lookup and Recipient Block List data sources?
  • How successful is the Sender Filter agent in evaluating and filtering out "suspect" senders?
  • Is the Sender ID agent efficient?
  • Has the Edge Transport server experienced latencies while connecting to the Exchange store? Which store interface can this delay be attributed to?
  • How many messages are available in the delivery queue? Is the number very high?
  • Do too many messages exist in the retry queue?
  • Are too many messages awaiting delivery to an external recipient?
  • Have messages been queued in the Unreachable queue?
  • Does the poison queue contain messages?
  • How efficient is Connection Filter agent? How many connection requests were rejected by the agent? Which data store used by the agent rejected the maximum requests - the IP Block list providers, IP allow list providers, or the IP Allow/Block lists defined by administrators?
  • Were any spams detected by the content filter agent?
  • Were any local/remote senders blocked by the Protocol Analysis agent?