eG Monitoring
 

Measures reported by LSRegistrarTest

Registrars are used to authenticate logon requests, and to maintain information about user status and availability. The Registrar is perhaps the most important component in Lync Server; without a Registrar, users would not be able to log on to the system, and Lync Server would not be able to keep track of users and their current status. When a user logs on to Lync Server, the endpoint the user is logging on from (be it a computer, a mobile phone, or some other device) sends a REGISTER request to the registration server; in turn the server responds by challenging the client device for authentication credentials. If the client passes the challenge (that is, if the client presents a valid set of credentials), then the user is authenticated and endpoint information such as IP address, port, and user name is logged in the registration database. When a user logs off, this information is then removed from the database. In between log on and log off, the Registrar keeps status information up-to-date and helps to route messages to and from the user. Often administrators may want to identify how well the information i.e., messages from the user are received and processed by the server and rate the user experience on the server. This is where the LSRegistrarTest test helps!

This test monitors the messages sent to the server and reports the following:

  • The number of queued messages that were dropped, the number of stored security associations and security records;
  • The number of incoming messages that timed out and those that were held for a longer duration;
  • The number of requests that were received by the registrar and the number of requests that were dropped;
  • The responses that were received for the incocming messages;
  • The DHCP requests and SIP protocol requests received etc.

The measures made by this test are as follows:

Measurement Description Measurement Unit Interpretation
Queue_Msg_notSent Indicates the number of messages in the message queue that could not be sent to the monitoring server and were eventually dropped. Number Ideally, the value of this measure should be zero. If the value of this measure is high, check if the message queuing is working properly. Also, ensure that the QoE Monitoring agent has permissions to access the MSMQ.
auth_sys_error Indicates the rate at which authentication failed due to system errors. Errors/Sec A high value indicates that this server may be experiencing problems when using Windows Authentication subsystem.
Total_SAs_Stored Indicates the total number of security associations that are currently stored in the server. Number  
StoredSecurity Indicates the total number of security records that are currently stored in the server. Number A security record is created for each user authenticated on the Lync server.
Holding_in_msg_time Indicates the average time the incoming messages that were being processed by the server were put on hold. Secs The value of this measure should be in the range of 1-3 seconds. The server will throttle new incoming messages after going above the high watermark and until the number of messages falls below the low watermark. The server starts rejecting new connections when the average holding time is greater than overload time of 15 seconds.

A high value for this measure indicates that the server is too busy and is unable to process the messages on time.
In_msg_timeout Indicates the number of incoming messages that timed out. Number The incoming messages may time out when the server is processing the messages beyond the maximum tracking interval.
InMsg_HighWM Indicates the number of incoming messages that were processed by the server were put on hold beyond the high watermark time threshold. Number A value close to 0 is desired for this measure.
InMsg_OverldWM Indicates the number of incoming messages that were processed by the server were put on hold beyond the overload watermark time threshold. Number  
AvgTime_InMsg Indicates the average time taken to process incoming SIP protocol messages. Secs  
IncomingMsg Indicates the rate at which the incoming SIP protocol messages are received. Messages/Sec  
EventsChgeRate Indicates the number of SIP transactions or dialog state change events that are currently processed. Number  
RequestRate Indicates the rate at which SIP protocol requests were received. Requests/Sec  
ReqDropRate Indicates the rate at which incoming SIP protocol requests were dropped. Requests/Sec The incoming requests/incoming responses may be dropped due to bad headers, insufficient routing information, server resource allocation failure etc.
ResDropRate Indicates the rate at which incoming SIP protocol responses were dropped. Requests/Sec
conn_refused_by_overload Indicates the number of connections that were refused due to the overload condition of the server. Number The connections may be refused with the Service Unavailable response when the server is overloaded.
Sockets Indicates the number of opened sockets. Number  
Res500_Rate Indicates the rate at which Local 500 Responses were generated by the server. Responses/Sec A Local 500 Response indicates that a server component of the Lync server is not functioning properly.
Resp503_Rate Indicates the rate at which Local 503 Responses were generated by the server. Responses/Sec A Local 503 Response corresponds to the server being unavailable. On a healthy server, you should not receive this code at a steady rate. However, during ramp up, after a server has been brought back online, there may be some 503 responses. Once all users get back in and the server returns to a stable state, there should no longer be any 503 responses returned.
Resp504_Rate Indicates the rate at which Local 504 Responses were generated by the server. Responses/Sec A Local 504 Response indicates connectivity issues with other servers. It can indicate connection failures or delays connecting to remote servers.
DHCPReq Indicates the total number of DHCP requests. Number  
ACKReq Indicates the total number of incoming SIP protocol ACK requests received. Number  
ReqRoute Indicates the number of SIP protocol requests that were routed using a routing table rule. Number  
Queue_Depth Indicates the average number of database requests that are currently waiting to be run. Number The value of this measure should be less than 500. A value above 500 indicates that the backend database server is busy and is currenly unable to process the requests.
Queue_Latency Indicates the average time spent by the database request in the database queue. Secs If for any reason resources (such as disk, memory, network, and processor) on the Back End Server are too high, the queue latency will also be high. Queue latency can also be high if there is too much network latency between the Front End Server and the Back End Server.
SprocLatency Indicates the average time taken by the backend database server to process an sproc call. Secs