SQLTRACE_INCREMENTAL_FLUSH_SLEEP

Description

Very similar to SQLTRACE_BUFFER_FLUSH except this wait accumulates between data flushes. SQL Server by default has at least one trace configured to capture configuration changes and other events that could be useful should problems be detected. Because of this wait will often be present but in low numbers. Generally, it is safe to ignore. However, if the values are high please see the suggested solutions below

Resolved by

DBAs

Suggested solutions

  1. Avoid running Profiler, use a scripted trace or Extended Events instead
  2. Do not run Profiler over the network
  3. When using Profiler or trace only capture the appropriate events and columns. Filter when possible
  4. Send trace output to file rather than to table
  5. Ensure that the trace destination file is to a fast disk

Additional research

Profiler Best Practices
Filtering Out Benign Waits

⇐ Back to index