Database Cache Size (MB)

Definition:
Shows the amount of system memory, used by the database cache manager to hold commonly used information from the database files to prevent file operations.

Information:
This and Database Cache Hit % are useful counters for gauging whether a server's performance problems might be resolved by adding more physical memory. Use this counter along with store private bytes to determine if there are store memory leaks.

Possible problems:
Situation when the database cache size seems too small for optimal performance and there is little available memory on the system (check the value of Memory/Available Bytes) could negatively impact on performance. If there is ample memory on the system and the database cache size is not growing beyond a certain point, the database cache size may be capped at an artificially low limit. Increasing this limit may increase performance.

Remediation:
Adding more memory to the system and/or increasing database cache size may increase performance.

N/A

Database Page Fault Stalls/sec

Definition:
Shows the rate that database file page requests require of the database cache manager to allocate a new page from the database cache.

Information:
Should be 0 at all times.

Possible problems:
If this value is nonzero, this indicates that the database is not able to flush dirty pages to the database file fast enough to make pages free for new page allocations.

Remediation:
If the disk subsystem is not meeting demand, correcting the problem may require additional disks, faster disks, or modifying the disk configuration.

> 0

Version Buckets Allocated

Definition:
Shows the total number of version buckets allocated.

Information:
Should be less than 12,000 at all times. The maximum default version is 16,384. If version buckets reach 70% of maximum, the server is at risk of running out of the version store.

Possible problems:
Typically indicates a database transaction which is taking a long time to save to disk. During Online DB defrags, the version buckets may increase.

Remediation:
Verify if the server has any applications running that have a long running transaction which has not been saved to disk, causing the version store memory resource to be exhausted.

> 11500

Log Record Stalls/sec

Definition:
Shows the number of log records that can't be added to the log buffers per second because the log buffers are full. If this counter is nonzero for a long period of time, the log buffer size may be a bottleneck.

Information:
The average value should be below 10 per second. Spikes (maximum values) should not be higher than 100 per second.

Possible problems:
If I/O log write latencies are high, check for RAID5 or synchronize replication on log devices.

Remediation:
You can also use the MSExchange Database Instances (Information store/<Database Name>)\log record stalls/sec counter to determine which database(s) may be having issues. This assists you in determining which drive(s) to focus on. This counter is an extended Exchange counter in Performance Monitor. The solution can include additional disks, reconfigured RAID configuration, adding new database(s), or re-balancing mailboxes across databases or servers.

> 10 for 15 polls

Log Threads Waiting

Definition:
Shows the number of threads waiting for their data to be written to the log to complete an update of the database. If this number is too high, the log may be a bottleneck.

Information:
Should be less than 10 on average.

Possible problems:
If this number is too high, the log may be a bottleneck. Regular spikes concurrent with log record stall spikes indicate that the transaction log disks are a bottleneck. If the value for log threads waiting is more than the spindles available for the logs, there is a bottleneck on the log disks.

Remediation:
If the disk subsystem is not meeting demand, correcting the problem may require additional disks, faster disks, or modifying the disk configuration.

> 10 for 15 polls