Foglight for Sybase 5.7.5.40 - User and Reference Guide

Investigating Stable Queues

Investigating Stable Queues

The Foglight Cartridge for Sybase assists your investigation into replication server stable queues with the RS_SQM_Trend table view, RS_SQM_Activity graph view, RS_SQM_QueueSize graph view, and the RS_SQM_TrailSize graph view.

Stable Queues are Store-and-forward queues where Replication Server stores messages destined for a route or database connection. Messages written into a stable queue remain there until they can be delivered to the destination Replication Server or database. Replication Server builds and allocates space for the stable queues from the disk partitions created for the Replication Server. The most referenced types of queues are inbound (transactions from a Primary Server) and outbound (transaction messages for another Replication Server via a route or to a direct Data Server connection - DSI).

Foglight allows a Replication Server administrator to monitor the overall performance of the stable queues over time by gathering the respective metrics and allowing these metrics to be viewed in pertinent graphs.

Monitoring the stable queues can be a measure of how much activity or volume is passing through the Replication Server. The Replication Server administrator can use this information to understand the performance profile of the monitored Replication Server. This information also serves as a tool for capacity planning.

The trending of SQM performance can be turned on or off via the Sybase_RS agent properties. Refer to the Sybase_MDA Agent Properties for this agent.

To start with the RS_SQM_Trend table view:

•	Open the RS_SQM_Trend table view to display activity totals of SQM threads.

To view more detailed information:

•	Drill down on the Messages Written or Messages Read line to view SQM messages read and written within the last polling period with the RS_SQM_Activity graph view.

TIP: This is a good indication of the volume of transactions that the Replication Server is handling.

•	Drill down on the Queue Size line to view SQM queue size with the RS_SQM_QueueSize graph view.

•

Drill down on the Data Not Processed line to view SQM queue size with the RS_SQM_TrailSize graph view. Use this to watch if the Stable Queue is keeping up with the incoming transactions. An increase in the size indicates the stable queue is falling behind in forwarding the transactions to their destination. Research downstream of the stable queue to find the cause.

Investigating Partition Status

Investigating Partition Status

The Foglight Cartridge for Sybase assists your investigation into partition status with the RS_PartitionStatusSummary graph view and the RS_PartitionStatusDetail table view.

Partitions in Replication Server are a raw disk partition or operating system file that Replication Server uses for stable queue storage.

Foglight monitors the Replication Server for any partition that is not in an online state. The RS_PartitionStatusSummary graph view provides a quick overview of partition status on the monitored Replication Server. Foglight returns more detailed information on any partition that is not online.

Foglight raises an alert if any partition is found in any state other than online.

Review the Replication Server error log for any additional information about an off-line partition. Check the location of the physical device and see if there are any reported errors on the physical volume.

To start with the RS_PartitionStatusSummary graph view:

•	Open this graph directly, or from the RS_Internal_Overview graph view after clicking Partitions Offline.

To view more detailed information:

•	Drill down on the Offline value to view detailed status information for partitions with the RS_PartitionStatusDetail table view.

Investigating Partition Space

Investigating Partition Space

The Foglight Cartridge for Sybase assists your investigation into partition space with the RS_PartitionSpaceUsedSummary table view and the RS_PartitionSpace_Trend table view. Partitions in Replication Server are a raw disk partition or operating system file that Replication Server uses for stable queue storage.

Foglight monitors the replication Server for overall Partition Space that is allocated to the stable queues. The RS_PartitionSpaceUsedSummary table view gives a quick overview of partition space on the monitored Replication Server.

The RS_PartitionSpace_Trend table view provides information to track individual partition space allocation over time. Trending must be turned on in order for this information to be available. Review the agent properties for the Sybase_RS Agent for more information on how to configure this.

To start with the RS_PartitionSpaceUsedSummary table view:

1

Open this view directly or from the RS_Internal_Overview graph view after clicking on Space Used Percentage. The RS_PartitionSpaceUsedSummary table view shows a breakdown of the space used by all partitions on the Replication Server. The Space percent free should be a good indicator when it may be needed to increase the amount of partition space needed by the Replication Server. If the Replication Server runs out of space to store messages, there is a negative impact on the Replication Server.

2	Use the “create partition” command to make more partition space available to the Replication Server. If the partition is to be an operating system file, then before using the “Create partition” command, you must first create the file (in UNIX use ‘touch’).

To view more detailed information:

•	Drill down on the Total Size (Mb) line to view detail information on each partition with the RS_PartitionSpace_Trend table view. The trending of Partitions space metrics can be turned on or off via the Sybase_MDA Agent Properties. Refer to the Setting Data Management Properties for this agent.

Investigating Error Log Messages

Investigating Error Log Messages

The Foglight Cartridge for Sybase assists your investigation into the status of the monitored Replication Server by returning pertinent messages Foglight finds in the Replication Server error log.

Replication Server errors are recorded in the Replication Server error log. A Replication Server error log contains informational, warning, thread-terminated, fatal, and internal error messages.

To start with the RS_ErrorlogSummary table view:

•	Open the RS_ErrorlogSummary table view to display summary information from the Replication Server error log. This view provides summary totals for each type of message in the error log found in the last poll cycle for the Sybase_RS Agent.

To view more detailed information:

•	Drill down on the Informational column to view detailed row information from the Replication Server error log with the RS_ErrorlogDetail table view.