Web proxy performance counters
Applies To: Forefront Threat Management Gateway (TMG)
The following table lists the performance counters for the Web Proxy.
| Performance Counter | Description |
|---|---|
Active Web Sessions |
Indicates how many clients are currently running Web Proxy. Monitoring this counter at both peak and off-peak times gives a good indication of server usage. The configuration setting for maximum Web request connections influences this value. This counter may also be useful if you need to temporarily stop Forefront TMG services. When authentication does not take place, all of the clients from a single IP address are viewed as one session. |
Array Bytes Received/sec (Enterprise) |
The rate at which data bytes are received from other Forefront TMG computers within the same array. |
Array Bytes Sent/sec (Enterprise) |
The rate at which data bytes are sent to other Forefront TMG computers within the same array. |
Array Bytes Total/sec (Enterprise) |
The sum of the Array Bytes Sent/sec and the Array Bytes Received/sec counters. This is the total rate for all data bytes transferred between the Forefront TMG computer and other members of the same array. |
Average Milliseconds/request |
The mean number of milliseconds required to service a Web Proxy client request, not including requests serviced by the Secure Sockets Layer (SSL) tunnel. This counter can be monitored at peak and off-peak times to get a comprehensive picture of the rate at which client requests are being serviced. A counter with a value that is too high might indicate that the Forefront TMG computer is having difficulty in handling all requests and that requests are being delayed. |
Average request speed |
The average speed of the request for all requests in the previous minute. The speed is calculated by dividing the request size by the time needed to serve the request, and is provided in bytes per second. |
Bytes actually requested from server for range requests/Bytes in range requests (%) |
The percentage requested from the server for range requests out of the number of bytes in range requests when the cache rule enabled range request caching. |
Bytes Requested from Server in Ranges |
The total number of bytes requested from the server in HTTP requests containing range headers. |
Bytes Served (Last Hour) From Cache in Ranges |
The total number of bytes returned in responses to HTTP requests containing range headers served from cache in the last hour. |
Bytes Served (Last Hour) in Ranges |
The total number of bytes returned in responses to HTTP requests containing range headers in the last hour. |
Bytes Served from Cache in Ranges |
The total number of bytes returned in responses to HTTP requests containing range headers served from cache. |
Bytes Served in Ranges |
The total number of bytes returned in responses to HTTP requests containing range headers. |
Cache Hit Percentage for Range Requests |
The percentage of bytes served from cache for responses to HTTP requests containing range headers. |
Cache Hit Ratio (%) |
Determines how many Web Proxy client requests have been served using cached data (Total Cache Fetches), as a percentage of the total number of successful Web Proxy client requests to the Forefront TMG computer (Total Successful Requests). Its value gives a good indication of the effectiveness of the cache. A high counter value indicates that a high level of requests are being serviced from the cache, meaning faster response times. A zero counter value indicates that caching is not enabled. A low counter value may indicate a configuration problem. The cache size may be too small, or requests may not be cacheable. |
Cache Hit Ratio for the Last 10K Requests (%) |
The amount of requests served from the cache as a percentage of total successful requests serviced. This ratio is the same as that measured by the Cache Hit Ratio counter. The difference between these two counters is that the Cache Running Hit Ratio counter measures this ratio for the last 10,000 requests serviced, and the Cache Hit Ratio counter measures this ratio since the last time that Web Proxy was started. This means that the Cache Running Hit Ratio counter gives a more dynamic evaluation of cache effectiveness. |
Client Bytes Received/sec |
The rate at which data bytes are received from Web Proxy clients. The value changes according to the volume of Web Proxy client requests, but a consistently slow rate may indicate a delay in servicing requests. |
Client Bytes Sent/sec |
The rate at which data bytes are sent to Web Proxy clients. The value changes according to the volume of Web Proxy client requests, but a consistently slow rate may indicate a delay in servicing requests. |
Client Bytes Total/sec |
The sum of the Client Bytes Sent/sec and the Client Bytes Received/sec counters. This is the total rate for all bytes transferred between the Forefront TMG computer and Web Proxy clients. |
Connect Errors |
The total number of errors that occurred while connecting. |
Connect Errors/Total Errors (%) |
The percentage of errors that occurred while connecting as a ratio of the total number of failed requests during the time shown in the duration field. |
Current Array Fetches Average (Milliseconds/request) |
The mean number of milliseconds required to service a Web Proxy client request that is fetched through another array member. This does not include requests for services by the SSL tunnel. |
Current Cache Fetches Average (Milliseconds/request) |
The mean number of milliseconds required to service a Web Proxy client request from the cache. This does not include requests for services by the SSL tunnel. |
Current Direct Fetches Average (Milliseconds/request) |
The mean number of milliseconds required to service a Web Proxy client request directly to the Web server or upstream proxy. This does not include requests for services by the SSL tunnel. |
Failing Requests/sec |
The rate of Web Proxy client requests that have been completed with some type of error. This counter can be compared with the Requests/sec counter to give an indication of how well Forefront TMG is servicing incoming Web requests. A high failure rate, as compared with the rate of incoming requests, suggests that Forefront TMG is having difficulty in coping with all incoming requests. Connection settings for incoming Web requests may be incorrectly configured, or connection bandwidth may be insufficient. |
Failing Requests/Total Requests (%) |
The percentage of failing requests out of the total number of requests during the time shown in the duration field. |
FTP Requests |
The number of File Transfer Protocol (FTP) requests that have been made to Web proxy. A consistently low counter value may influence the caching policy for FTP objects. |
HTTP Requests |
The number of HTTP requests that have been made to Web proxy. |
Incoming Connections/sec |
The rate of incoming connections (per second). |
IO Errors to Array Member |
The total number of input/output errors that occurred during communication with an array member. |
IO Errors to Array Member/Total Errors (%) |
The percentage of input/output errors that occurred during communication with an array member out of the total number of failed requests during the time shown in the duration field. |
IO Errors to Client |
The total number of input/output errors that occurred during communication with the client. |
IO Errors to Client/Total Errors |
The percentage of input/output errors that occurred during communication with the client out of the total number of failed requests during the time shown in the duration field. |
IO Errors to Server |
The total number of input/output errors that occurred during communication with the server. |
IO Errors to Server/Total Errors (%) |
The percentage of input/output errors that occurred during communication with the server out of the total number of failed requests during the time shown in the duration field. |
Maximum Users |
The maximum number of users that have connected to Web proxy simultaneously. This counter can be useful for determining load usage and license requirements. |
Memory Pool for HTTP Requests (%) |
Percentage of memory available for HTTP requests. When an HTTP request is made, Forefront TMG uses memory from a preallocated pool. You can use the ProxyVmemAlloc3pSize registry value in the HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\W3Proxy\Parameters registry key to modify the size of this pool. |
Memory Pool for SSL Requests (%) |
Percentage of memory available for SSL requests. When an SSL request is made, Forefront TMG uses memory from a preallocated pool. You can use the ProxyVmemAlloc1pSize registry value in the HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\W3Proxy\Parameters registry key to modify the size of this pool. |
Outgoing Connections/sec |
The rate of outgoing connections (per second). |
Protocol anomalies blocked by NIS in last day |
The total number of protocol anomalies which the Network Inspection System blocked in last day. |
Protocol anomalies detected by NIS in last day |
The total number of protocol anomalies which the Network Inspection System detected in last day. |
Requests from Array Member (Enterprise) |
The total number of requests coming from another array member. |
Requests from Array Member/Total Errors (%)(Enterprise) |
The percentage of requests coming from another array member out of the total number of failed requests during the time shown in the duration field. |
Requests to Array Member (Enterprise) |
The total number of requests going to another array member. |
Requests to Array Member/Total Errors (%)(Enterprise) |
The percentage of requests going to another array member out of the total number of failed requests during the time shown in the duration field. |
Requests with Keep Alive to Array Member |
The total number of requests that use an existing keep-alive connection between Forefront TMG and another array member. |
Requests with Keep Alive to Array Member/Total Errors (%) |
The percentage of requests that use an existing keep-alive connection between Forefront TMG and another array member out of the total number of failed requests during the time shown in the duration field. |
Requests with Keep Alive to Client |
The total number of requests that use an existing keep-alive connection between the client and Forefront TMG. |
Requests with Keep Alive to Client/Total Errors (%) |
The percentage of requests that use an existing keep-alive connection between the client and Forefront TMG as a ratio of the total number of failed requests during the time shown in the duration field. |
Requests with Keep Alive to Server |
The total number of requests that use an existing keep-alive connection between Forefront TMG and the Web server. |
Requests with Keep Alive to Server/Total Errors (%) |
The percentage of requests that use an existing keep-alive connection between Forefront TMG and the Web server as a ratio of the total number of failed requests during the time shown in the duration field. |
Requests with Multiple Ranges |
The number of requests with more than one range in the HTTP Range header. |
Requests/sec |
The rate of incoming requests that have been made to Web proxy. A higher value means that more Forefront TMG resources will be required to service incoming requests. |
Reverse Bytes Received/sec |
The rate at which data bytes are received by Web proxy from Web publishing servers in response to incoming requests. This rate can be monitored at peak and off-peak times as an indication of how Forefront TMG is performing in servicing incoming Web requests. |
Reverse Bytes Sent/sec |
The rate at which data bytes are sent by Web proxy to Web publishing servers in response to incoming requests. This rate can be monitored at peak and off-peak times as an indication of how Forefront TMG is performing in servicing incoming Web requests. |
Reverse Bytes Total/sec |
The total sum of the Reverse Bytes Sent/sec and the Reverse Bytes Received/sec counters. This is the total rate for all bytes transferred between Web proxy and Web publishing servers in response to incoming requests. |
Signatures blocked by NIS in last day |
The total number of signatures which the Network Inspection System blocked in last day. |
Signatures detected by NIS in last day |
The total number of signatures which the Network Inspection System detected in last day. |
Sites Allowed |
The total number of Internet sites to which Web proxy has denied access. An excessively high number might indicate an access policy that is too restrictive. |
Sites Allowed in last day |
The total number of Web sites to which the Web Proxy filter allowed access in last day. |
Sites Denied |
The total number of Internet sites to which Web proxy has granted access. This can be compared with the Site Access Denied counter to give a numeric summary of the results of access policy configuration. |
Sites Denied in last day |
The total number of Web sites to which the Web Proxy filter denied access in last day. |
SNEWS Sessions |
The total number of SNEWS sessions serviced by the SSL tunnel. |
SSL Client Bytes Received/sec |
The rate at which SSL data bytes are received by Web proxy from SSL clients. Similar to the Client Bytes Received/sec counter, but counts only SSL requests. |
SSL Client Bytes Sent/sec |
The rate at which SSL data bytes are sent by Web proxy to SSL clients. Similar to the Client Bytes Sent/sec counter, but counts only SSL requests. |
SSL Client Bytes Total/sec |
The sum of the SSL Client Bytes Sent/sec and the SSL Client Bytes Received/sec counters. This is the total rate for all bytes transferred between Web proxy and SSL clients. |
Thread Pool Active Sessions |
The number of sessions being actively serviced by thread pools. |
Thread Pool Failures |
The number of requests rejected because the thread pool was full. |
Thread Pool Size |
The number of threads in the thread pool. This thread pool represents the resources available to service client requests. |
Total Array Fetches (Enterprise) |
The total number of Web Proxy client requests that have been served by requesting the data from another Forefront TMG computer within this array. These requests are the result of the Cache Array Routing Protocol (CARP) algorithm, which randomly stores objects in any one of the member servers cache. This counter is influenced by the cache size for each Forefront TMG computer in the array, because a server with a larger cache holds more cache items. The load factor for each server can also be configured to determine how workload is divided among array members. |
Total Cache Fetches |
The total number of Web Proxy client requests that have been served by using cached data. A high number indicates a cache being fully exploited. |
Total Failing Requests |
The total number of requests that have failed to be processed by Web proxy due to errors. Errors can be the result of Web proxy failing to locate a requested server URL on the Internet or because the client did not have authorized access to the requested URL. This counter should be lower than the Total Successful Requests counter. If not, it is an indication that Forefront TMG is failing to service requests effectively. This could be a configuration problem, or a connection that is too slow. It could also indicate an access policy that is too restrictive. |
Total Pending Connects |
The total number of pending connections to Web proxy. |
Total Requests |
The total number of requests that have been made to Web proxy. It is the sum of the Total Successful Requests and the Total Failed Requests counters. |
Total Reverse Fetches |
The total number of incoming requests that have been served by requesting the data from Web publishing servers. |
Total SSL Sessions |
The total number of SSL sessions serviced by the SSL tunnel. |
Total Successful Requests |
The total number of requests that have been successfully processed by Web proxy. This counter can be compared with the Total Requests and the Total Failed Requests counters to indicate the effectiveness of Forefront TMG in servicing requests. |
Total Upstream Fetches |
The total number of requests that have been served by using data from the Internet or from a chained proxy computer. This counter can be compared to the Total Cache Fetches counter to determine what proportion of requests are being serviced from remote servers on the Internet or upstream proxies, compared with those being serviced from the cache. |
Total Users |
The total number of users that have connected to Web proxy. It represents a history of past server usage. |
Unknown SSL Sessions |
The total number of unknown SSL sessions serviced by the SSL tunnel. |
Upstream Bytes Received/sec |
The rate at which data bytes are received by Web proxy from remote servers on the Internet or from a chained proxy computer in response to requests from Web proxy. The value of this counter depends partially on the connection bandwidth. If the counter value is consistently low, it may indicate a bottleneck caused by a slow connection. Changing the bandwidth priority configuration may help in this situation, or a faster connection may be required. |
Upstream Bytes Sent/sec |
The rate at which data bytes are sent by Web proxy to remote servers on the Internet or to a chained proxy computer. The value of this counter depends partially on the connection bandwidth. If the counter value is consistently low, it may indicate a bottleneck caused by a slow connection. Changing the bandwidth priority configuration may help in this situation, or a faster connection may be required. |
Upstream Bytes Total/sec |
The sum of the Upstream Bytes Sent/sec and the Upstream Bytes Received/sec counters. It represents the total rate for all bytes transferred between Web Proxy and remote servers on the Internet or a chained proxy server. |