Monitoring Microsoft Skype for Business

To avoid this, the eG Enterprise monitors the Microsoft Skype for Business Server and captures anomalies before users notice and complain.

Layer model

Figure 1 : The layer model of the Microsoft Skype for Business

Using the metrics reported, administrators can find quick and accurate answers for the following performance questions:

  • Observe address book accesses, measure the speed of these accesses, and report slowdowns;
  • Monitor the message processing ability of the server and report deficiencies;
  • Capture failed messages and when failures occurred – during message validation? in the MSMQ queue? or when written to the database?
  • Measure the quality of the A/V conference experience with the Skype for Business server;
  • Understand how well the server processes call park requests and report slowdowns;
  • Know how many users/clients are currently connected to the server, and thus gauge the current load on the server; 
  • Identify dropped conference activities and unfinished tasks
  • Monitor the server’s interactions with the database and in the process, reveal the requests queued for processing by the database and the time these requests spent in queue; database slowdowns can thus be captured; 
  • Rapidly detect client and server authentication failures and DNS resolution failures;
  • Messages that could not be sent to the server
  • DNS resolution failures
  • The current health and draining state of the Application sharing Conference unit, the Data MCU server and the Instant Messaging multipoint control unit;
  • Know how well the conferences are handled by the Application sharing conference unit
  • How many RDP connections failed in the Application sharing Control unit?
  • How many whiteboards were served by the Data MCU server and how many conferences are currently active on the Data MCU server?;
  • How many conferences were active on the Instant Messaging multipoint control unit and how many message delivery failure notification messages were sent from the server?;
  • Know how many add user and add conference responses failed in the server
  • How many conferences were initiated from each service and how many times conference processing was actiually delayed?;
  • How many times the application endpoint creation failed for the Conferencing Auto Application service and how many incoming calls were actually incomplete?
  • Monitor the message processing ability of the server and report deficiencies;
  • How many SIP connections failed and how many are currently active?;
  • How long the distribution list take to process the requests?
  • How many times the Join launcher service failed and how many incoming requests were received by the service?
  • How many replication requests were received by the replication service and how long does the service take to process?;
  • Continuously monitor the mobility services and figure out the successful Get Location requests, the push notifications that failed and throttled;
  • How many times the user provisioning and publish calls failed?
  • Monitor the backend database server and figure out how well requests are processed and how long does it take to process the requests?;
  • How well the stored procedure calls were executed by the user service?;
  • How many stale items were in the queue of the Luync storage service and how much space did the service utilize to store the items to the backend database server?;
  • For each client version, how many unique users were actually connected?;
  • How many users were actually connected to the server?
  • For how many users were voice calls enabled?
  • How many times the HTTPS connections failed?

The Operating System, Network, TCP and Application Process layers of the Microsoft Skype for Business monitoring model is similar to that of a Windows Generic server model. The tests associated with the above-mentioned layers have been discussed extensively in Unix and Windows Server model.