Microsoft SQL Server Distributed Replay (Troubleshooting) Part 6 :

 

While running the Distributed Replay i ran into couple of issues listed below.

 

How long the trace should be left running?

Running trace should be well planned with Support/Dev/QA team to generate the peak load which is used to set the benchmark for the application/website.The trace should include heavy insert/update/delete/Long running queries/business reports.

Try to capture trace events for less then an hour and do not forget to set database as filter.

 

Trace replays quickly with lot of failed events?

2015-06-06 09:18:29:899 Info DReplay    Event replay in progress.
2015-06-06 09:18:49:915 Info DReplay    Client1: 4529 events replayed, 179 events succeeded (pass rate 3.95 %).
2015-06-06 09:18:49:915 Info DReplay    Client2: 3137 events replayed, 491 events succeeded (pass rate 15.65 %).

This indicates that 4529 – 179= 4350 events where failed without completing the transaction/batch.This could also happen due to permissions. The executer should have alter trace permission on the target SQL Server instance.

 

How to change the Controller Name configured for client at the time on installation?

Navigate to C:Program Files (x86)Microsoft SQL Server120ToolsDReplayClient and open DReplayClient.config file.

image

How and where error are captured?

Dreplay process creates ReplayResult.trc @ C:Program Files (x86)Microsoft SQL Server120ToolsDReplayClientResultDir.Open it profiler tool.

Leave a Reply

Your email address will not be published. Required fields are marked *