Operations Manager (OpsMgr) 2007 collects a number of performance counters for the agent (monitored computer), management server, and Audit Collection System (ACS) collector. These counters are created when the corresponding component is installed on a computer.
Each managed agent maintains two sets of performance counters. Management servers maintain these counters as well.
Health Service counters—. These are performance counters for the Health Service running on that computer (see Table B.1).
Table B.1. Health Service Counters
Counter[*]
Description
Malformed Data Items Dropped
The total number of malformed data items that have been dropped
Malformed Data Items Dropped Rate
The rate at which malformed data items are being dropped
Module Count
The number of modules running in the Health Service
Task Count
The number of tasks running in the Health Service
Workflow Count
The number of workflows running in the Health Service
[*] There will be an instance of each of these per management group.
Health Service Management Group counters—. These are performance counters relating to activity for the management group associated with that agent (see Table B.2). If the agent reports to multiple management groups (multihomed), there will be multiple instances of the Health Service Management Group counters.
Table B.2. Health Service Management Group Counters
Counter[*]
Description
Active File Downloads
The number of file downloads to this management group that are currently active.
Active File Uploads
The number of file uploads from this management group that are currently active.
Active Batch Processing Time
The average amount of time in milliseconds it takes to process a batch of data.
Average Bind Write Action Batch Size
The average amount of time in milliseconds it takes to process a batch of data.
Batch Processing Rate
The rate at which batches of data are being processed.
Bind Data Source Average Batch Size
The average number of items inserted per batch into the workflow which originated from a remote Health Service.
Bind Data Source Batch Post Rate
The rate at which batches that originated from a remote Health Service inserted into the workflow.
Bind Data Source Item Drop Rate
The rate at which items are dropped by the bind data source because the workflow cannot process them quickly enough.
Bind Data Source Item Incoming Rate
The rate at which items are received from another Health Service.
Bind Data Source Item Post Rate
The rate at which items received from another Health Service are inserted into the workflow.
Bind Write Action Batch Insertion Rate
The rate at which batches are sent through the workflow destined for a remote Health Service.
Bind Write Action Multiple Item Batches
The number of batches sent through the workflow destined for a remote Health Service with more than one item in them.
Bind Write Action Single Item Batches
The number of batches sent through the workflow destined for a remote Health Service that have exactly one item in them.
Bind Write Action Zero Item Batches
The number of batches sent through the workflow destined for a remote Health Service that have zero items in them (NULL Sets).
Incoming Management Data Rate
The rate at which management data is received by the Health Service.
Item Incoming Rate
The rate at which items are being received by the Health Service.
Maximum Send Queue Size
The maximum size of the Send Queue.
Send Queue % Used
The percentage of the Send Queue in use.
Send Queue Ack Timeout Rate (>4 send attempt)
The number of Ack timeouts for items that have been sent more than four times.
Send Queue Ack Timeout Rate (1 send attempt)
The number of Ack timeouts for items that have been sent only once.
Send Queue Ack Timeout Rate (2 send attempts)
The number of Ack timeouts for items that have been sent twice.
Send Queue Ack Timeout Rate (3 send attempts)
The number of Ack timeouts for items that have been sent three times.
Send Queue Ack Timeout Rate (all send attempts)
The rate of items scheduled to re-send because they have not been acknowledged and are marked as requiring guaranteed delivery.
Send Queue Incoming Acknowledgement Rate
The rate of incoming acknowledgements for items sent with guaranteed delivery.
Send Queue Insertion Rate
The number of items inserted into the queue.
Send Queue Insertion Rate (Guaranteed Delivery)
The number of items marked as guaranteed delivery inserted into the queue.
Send Queue Insertion Rate (Non-Guaranteed Delivery)
The number of items marked as nonguaranteed delivery inserted into the queue.
Send Queue Item Expiration Rate
The rate at which items are being purged from the queue because they have not been sent within the queue expiration time.
Send Queue Send Rate
The number of items sent from the queue.
Send Queue Send Rate (Guaranteed Delivery)
The number of items sent that are marked for guaranteed delivery. This only records items sent; it does not account for whether or not the items are acknowledged.
Send Queue Send Rate (Non-Guaranteed Delivery)
The number of items sent that are not marked for guaranteed delivery.
Send Queue Size
The size of the send queue in bytes, not including any journaled data.
[*] There will be an instance of each of these per management group.
In addition to the counters referenced in Tables B.1 and B.2, all management servers maintain performance counters for a number of items.
The first three groups of performance counters are located on each management server, but the associated service only runs when that management server is a Root Management Server (RMS). When the server is not acting as the RMS, these counters do not collect data.
OpsMgr Config Service (see Table B.3)
Table B.3. OpsMgr Config Service
Counter[*]
Description
Avg Management Pack Request
Average time to service a management pack (MP) request
Avg Request Queue Length
Average length of the request queue
Avg Request Queue Wait Time
Average wait time of requests in the queue
Change Notifications Failed
Number of change notifications that failed
Dirty State Notifications Failed
Number of dirty state notifications that could not be sent
Management Pack Requests Failed
Number of MP requests that could not be processed
Number of Active Requests
Number of requests being processed
Number of Queued Requests
Number of requests waiting in the queue
Number of Requests per Sec
Requests competed per second
[*] These are active on the RMS and are per management group.
OpsMgr Config Service per Agent (see Table B.4)
Table B.4. OpsMgr Config Service per Agent
Counter[*]
Description
Avg. State Sync Request (sec)
Average time to service a State Sync Request from this agent
State Sync Requests Failed
Number of State Sync Requests received from this agent that could not be processed
State Sync Requests Received
Number of State Sync Requests received from this agent
[*] These are active on the RMS and are per management group.
OpsMgr SDK Service (see Table B.5)
Table B.5. OpsMgr SDK Service
Counter[*]
Description
Client Connections
Number of current clients connected to the OpsMgr SDK service
Client Connections Using Cache
Number of current clients connected to the OpsMgr SDK service that are using the cache
Pending Client Type Cache Refresh
Number of currently executing type cache refresh threads in the OpsMgr SDK service
[*] These are active on the RMS and are per management group.
The next sets of counters are maintained on each management server.
OpsMgr DB Write Action Cache (see Table B.6)
OpsMgr DB Write Action Modules (see Table B.7)
Table B.7. OpsMgr Write Action Modules
Counter[*]
Description
Avg. Batch Size
Average number of entries processed by the module
Avg. Processing Time
Amount of time it takes to process an incoming batch on average
[*] These occur on each management server for the following instances: Alert Write Module, Discover Write Module, Event Write Module, Performance Signature Write Module, Performance Write Module, SQLobj Write Module, and State Change Module.
OpsMgr DW Synchronization Module (see Table B.8)
Table B.8. OpsMgr DW Synchronization Module
Counter[*]
Description
Avg Batch Processing Time, ms.
Average batch processing time in milliseconds
Avg Batch Size
Average batch size
Batch Age/sec
Batch age in seconds
Batch Size
Batch size
Batches/sec
Batches per second
Data Items/sec
Data items per second
Errors/sec
Errors per second
Total Error Count
Total error count
[*] These occur on each management server for the following instances: Configuration, Domain Snapshot, Health Service Outage, Maintenance Mode, Managed Identity, Relationship, and Type Managed Entity.
OpsMgr DW Write Module (see Table B.9)
Table B.9. OpsMgr DW Writer Module
Counter[*]
Description
Avg. Batch Processing Time, ms
Average batch processing time in milliseconds
Avg Batch Size
Average batch size
Batches/sec
Batches per second
Data Items/sec
Data items per second
Dropped Batch Count
Dropped batch count
Dropped Data Item Count
Dropped data item count
Errors/sec
Errors per second
Total Error Count
Total error count
[*] These occur on each management server for the Alert, Event, Performance, and State instances.
The OpsMgr Connector is located on management servers and database servers. The Connector is an interface to the OpsMgr SDK Service that allows applications to communicate synchronously with the RMS. Counters used by the OpsMgr Connector are listed in Table B.10.
Table B.10. OpsMgr Connector Counters
Counter | Description |
|---|---|
Bytes Decrypted | The number of bytes decrypted by the connector per second. This may be more or less than the number of data bytes received by the connector due to compression. |
Bytes Encrypted | The number of bytes encrypted by the connector per second. This may be more or less than the number of data bytes submitted for transmission due to compression. |
Bytes Received | The total number of network bytes received by the OpsMgr connector per second. This may be more or less than the number of data bytes received due to compression and encryption. |
Bytes Transmitted | The total number of network bytes transmitted by the OpsMgr connector per second. This may be more or less than the number of data bytes submitted for transmission due to compression and encryption. |
Fragmented Compression Packets | The number of compression packets that were received by the connector and had to be reassembled per second. This can be caused by a number of factors, including the size of the packet, the Maximum Transmission Unit (MTU) of the network interface, the quantity of traffic on the network, and/or the amount of load on the connector. Packet reassembly can hurt performance and the ability of the server to scale. |
Fragmented Data Packets | The number of data packets that were received by the connector and had to be reassembled per second. This can be caused by a number of factors, including the size of the packet, the MTU of the network interface, the quantity of traffic on the network, and/or the amount of load on the connector. Packet reassembly can hurt performance and the ability of the server to scale. |
Fragmented Encryption Packets | The number of encrypted packets that were received by the connector and had to be reassembled per second. This can be caused by factors including the size of the packet, the MTU of the network interface, the quantity of traffic on the network, the encryption algorithm used, and/or the amount of load on the connector. Packet reassembly can hurt performance and the ability of the server to scale. |
Fragmented Session Packets | The number of session packets that were received by the connector and had to be reassembled per second. This can be caused by a number of factors, including the size of the packet, the MTU of the network interface, the quantity of traffic on the network, and/or the amount of load on the connector. Packet reassembly can hurt performance and the ability of the server to scale. |
Fragmented SSPI Packets | The number of Security Support Provider Interface (SSPI) negotiation packets that were received by the connector and had to be reassembled per second. This can be caused by a number of factors, including the size of the packet, the MTU of the network interface, the quantity of traffic on the network, and/or the amount of load on the connector. Packet reassembly can hurt performance and the ability of the server to scale. |
I/O Errors | The number of I/O errors encountered by the connector per second. |
I/O Operations Timed Out | The number of I/O operations on the connector that were timed out due to inactivity per second. |
Open Connections | The number of TCP/IP (Transmission Control Protocol/Internet Protocol) connections currently open in the connector. |
Server Listen Queue Length | The number of outstanding listen requests that are queued. If this drops to zero, the connector will not be able to accept incoming connections. |
Table B.11 lists the counters used the OpsMgr Web Console application.
On the ACS Collector, we find two specific performance objects:
ACS Collector (see Table B.12)
Table B.12. ACS Collector Counters
Counter
Description
Connected Clients
Number of clients currently connected.
Database Queue % Full
This is the ratio of the number of events currently in the database loader queue divided by the size of the database loader queue, expressed as a percentage.
Database Queue Backoff Threshold
The collector will not allow new connections while the database queue length is greater than this threshold.
Database Queue Disconnect Threshold
The collector will disconnect existing connections while the database queue length is greater than this threshold.
Database Queue Length
This is the number of events currently in the database loader queue.
DB Loader Event Inserts/sec
This is the number of records per second inserted into the dtEvent database table.
DB Loaded Principal Inserts/sec
This is the number of records per second inserted into the dtPrincipal database table.
DB Loader String Inserts/sec
This is the number of records per second inserted into the dtString database table.
DB Principal Cache Hit %
This is the percentage of all principal handling requests handled by the string cache, avoiding a principal insert or lookup in the database.
DB Request Queue Length
This is the number of requests from the collector currently waiting to be serviced by the ACS database. These requests are used during forwarder handshake and database maintenance and are not part of normal event handling.
DB String Cache Hit %
This is the percentage of all string-handling requests handled by the string cache, avoiding a string insert or lookup in the database.
Event time in collector in milliseconds
This is the current time between event arrival at the collector and insertion in the database queues in milliseconds.
Incoming Events/sec
This is the current time between event arrival at the collector and insertion in the database queues in milliseconds.
Interface Audit Insertions/sec
Number of interface audit insertions per second.
Interface Queue Length
Length of the internal queue to the subscriber interface.
Registered Queries
Number of queries currently registered.
ACS Collector Client (see Table B.13)
Table B.13. ACS Collector Client Counters
Counter[*]
Description
Average Time between event generation and collection and seconds
This is the average time between event generation time (creation timestamp) and collection time in milliseconds.
Incoming Audit Size
This is the total size of events per second arriving at the collector from a specific forwarder.
Incoming Audits/sec
This is the total number of events per second arriving at the collector from a specific forwarder.
[*] These occur for each collector/client pair.