ARR Events

  • Article

Applies To: Windows 7, Windows Server 2008, Windows Server 2008 R2, Windows Vista

This topic provides you with detailed information about Application Request Routing (ARR) events that you can use to troubleshoot the servers in your server farm. For more information about ARR, see Application Request Routing and ARR Key Concepts and Features.

Warning Events

Event ID: 1000

Server <server name> in server farm <server farm name> is being marked unhealthy due to <reason>.

ARR uses two different methods to determine the health of the servers in a server farm:

  • Leverages the live traffic.

  • Uses a specific URL for health testing.

When either of these tests fails, ARR marks the corresponding server(s) as Unhealthy so that requests cannot be forwarded to the unhealthy server(s). This is not a fatal condition because an otherwise healthy server may have been taken offline for maintenance. ARR will continue to function normally and end users will not notice any failures as long as there are a sufficient number of healthy servers to handle the overall traffic load. However, if all servers are marked Unhealthy, end users will receive a 502.3 Bad Gateway error message.

Event ID: 1002

Server farm <name> has reached its minimum server threshold. No more servers will be marked unhealthy until the number of healthy servers increase.

ARR uses a concept of minimum servers where ARR expects to have at least the specified number of healthy servers in a server farm. For example, a server farm has 10 servers and the administrator has specified that at least five servers must be healthy in order to maintain the normal flow of traffic. When the number of healthy servers falls below the specified number of minimum servers, this event is logged and ARR disregards the health status of the servers. This can potentially cause some requests to be forwarded to unhealthy servers, causing further damage to the remaining healthy servers. This is by design. Although it is not ideal, this approach affects only some end users, rather than all end users if the healthy servers are overloaded.

Event ID: 1004

There was an error initializing the external cache for storing host-affinity. The data contains the error code.

This event is logged if there are problems using the External Cache module with ARR. Possible causes for this error include an incorrect path location to the External Cache file share or incorrect credentials to access the External Cache file share.

Event ID: 1005

There was an error while scavenging the content in disk location <file path>. <Number> files containing <number> bytes were deleted and <number> files containing <number> bytes remaining. The data contains the error code.

The disk scavenging task can fail while running due to several reasons. This event is logged if a failure occurs; the event provides more a more detailed error code for further investigation.

Event ID: 1006

Drive with path <drive location> is being marked unhealthy. The data contains the error code.

When there is an I/O error, ARR marks the corresponding disk as Unhealthy so that the disk can no longer be used for disk caching.

Informational Events

Event ID: 1001

Server <server name> in server farm <server farm name> is being marked healthy.

ARR uses two different methods to determine the health of the servers in a server farm:

  • Leverages the live traffic.

  • Uses a specific URL for health testing.

Both tests can be used to mark a healthy server as an unhealthy server; however, only the URL test can be used to change an unhealthy server back to a healthy status. This strategy is used because ARR does not sacrifice live traffic and knowingly sends the traffic to an unhealthy server. This event is logged when the URL test determines that an unhealthy server is now healthy, for example because the maintenance on the server is complete.

Event ID: 1003

Server <server name> in server farm <server farm name> changed state from <status> to <status>.

A user who has administrator user rights can mark the health status of a server in a server farm as Healthy or Unhealthy. This event is logged when the user takes this action. The event is informational because the action is intentional.

Event ID: 1007

Drive with path <drive path> is being marked healthy.

Due to an I/O error, ARR might mark a disk location as Unhealthy. When disk scavenging runs, ARR scans the healthy and unhealthy disk locations to determine whether the statuses are still accurate. If a disk location that is marked Unhealthy is now Healthy, disk scavenging will update the status so that ARR can use the disk again. This situation can occur if a user who has administrator user rights swaps an unhealthy drive with a healthy drive.

Event ID: 1008

Scavenger successfully ran for content in disk location <path>. <Number> files containing <number> bytes were deleted and <number> files containing <number> bytes remaining.

Disk scavenging finishes when disk usage reaches the low threshold. The low threshold indicates the percentage of configured disk space that will remain after files are deleted. When disk scavenging finishes, this event is logged. The log will include additional information such as the number of files and the amount of data that was removed.

Event ID: 1009

Scavenger started for content in disk location <drive path>.

Disk scavenging starts when disk usage reaches the high threshold. The high threshold indicates the maximum percentage of configured disk space that can be reached and that will be available for cached content before files are deleted. When disk scavenging starts, both the start and the finish events are logged so that information is available about the time when disk scavenging started running and the duration of time that it has been running.