- Pinpoints the SQL statements that do not complete on a timely basis due to resource contention or were queued due to a lack of available resources.
- Wait condition wait times are tracked in real-time by SQL statement, stored procedure, wait type, application, database and the end-user.
- View historical SQL wait time history by year, month, week, day, hour and drill down to a one second view of SQL wait times, I/O and CPU usage.
- Displays in one convenient graph SQL wait time, I/O, CPU usage, SQL text and SQL plan performance issues along with any subset of corresponding performance counters.
- Wait conditions and how to eliminate them are explained from our knowledge base of the 600+ wait conditions.
- STEALTH wait time monitoring only occurs when SQL runs longer than N seconds. Requires the Zero Impact Agent SQL Capture Agent. Very low overhead on high volume servers.
- Reducing SQL wait time -> reduces end-user response time -> improves server throughput -> no unnecessary hardware upgrade.
- Database and operating system performance counter monitoring with real-time alerting. View in real-time or over any prior time period.
- SQL statement I/O monitoring with automated plan mining. Pinpoints SQL having high I/O, CPU usage, long run times and poor plans.
- Database data and log file I/O stall/wait time monitoring monitors the I/O wait time for all database devices/files.
- TempDb contention monitoring. Pinpoints SQL experiencing wait times on the TempDB database.
- Monitors blocking and deadlocks.
- Database and disk drive space monitoring. View the entire Server Farm database and disk drive space usage in one 3D graph.
- Procedure cache and buffer cache monitoring. View the entire Server Farm procedure cache or buffer cache usage in one 3D graph.
- Baseline analysis identifies wait types, wait times by database device, SQL response time, performance counters and SQL jobs that have degraded or improved performance.
- Daily baseline analysis of the day's performance is compared in 3D graphs to prior time periods.
- SQL statements, stored procedures, wait types, applications, databases and end-users that have increasing wait times (and increasing end-user response time) are immediately identified.
- Allows performance issues to be proactively resolved before end-users are impacted.
- Tracks the performance of every unique SQL request.
- Ad hoc comparison of end-user response time, SQL performance, wait times, I/O bottlenecks, performance counters and job run times between servers and time periods.
- FREE Zero Impact SQL Capture Agent with every license to use when needed. Immediately pinpoints all poor performing production SQL.
- Optionally enable disable at any time.
- Non-intrusively sniffs the SQL packet flow. Does NOT connect to monitored server NOR impact the network.
- Captures all SQL text, bind parameters, logins, errors, end-user response time, network time and more.
- View the detailed SQL activity over any time period.
- Does NOT use ineffective SQL sampling or intrusive database server SQL traces.
- Top N Sql Analysis of 100% of the SQL activity.
- Enable the agent on priority servers where end-user response time is important, on servers experiencing performance issues that cannot be pinpointed with agentless monitoring.
- Allows accurate comparision of performance before and after application turnovers, release upgrades, hardware changes, indexing changes and SQL code changes.
- Step 1a. Zero Impact Sql Capture Agent captures 100% of the current production server SQL activity for a peak business day. Captures all SQL text, bind parameters, logins, end-user response time, network time, rows returned and more. Non-intrusively sniffs the SQL packet flow. Agent does NOT connect to the monitored server NOR impact the network. Has a low 1% overhead. No intrusive SQL traces are used.
- Step 1b. Our Agentless Bottleneck Monitor concurrently captures the current production environment wait conditions with their wait times, I/O device stall/wait times, TempDb contention, blocking, deadlocks, poor SQL plans and performance counters.
- Step 2. Next perform a multi-threaded replay of the captured SQL activity against the new SQL Server 2008 or 2012 release.
- Step 3a. Zero Impact Sql Capture Agent again captures 100% the replayed SQL activity with its performance during the multi-threaded replay. The end-user response time of every client replayed SQL request is captured.
- Step 3b. Agentless Bottleneck Monitor again concurrently captures the new release’s wait times, device stall/wait times, TempDb contention, blocking, deadlocks, poor SQL plans and performance counters. Wait times, I/O and CPU time are captured by SQL statement, stored procedure, wait type, application, database and end-user.
- Final step. Analysis of the multi-threaded SQL replay of results. the end-user response time of every unique SQL request. The SQL that ran faster or slower is identified. The percentage increase or decrease in response time by unique SQL request is now known.
- Wait times, I/O and CPU time are compared by SQL statement, stored procedure, wait type, application, database and end-user. Both decreased and increased wait times may be viewed for the day, by the hour, by the minute and down to a second.