dfc trace performance anaysis

1)设置dfc.properties enable dfc trace
dfc.tracing.enable=true
dfc.tracing.verbose=true
dfc.tracing.max_stack_depth=0
dfc.tracing.include_rpcs=true
dfc.tracing.mode=compact
dfc.tracing.include_session_id=true

2)获取trace 文件，使用awk进行转换，使trace可读性更好，有利于分析

3)traceD6.awk

This script converts a DFC trace file to a file ready to be loaded into Microsoft Excel in order to identify big/expensive queries that take a long time to return.

- Execute the following command

    awk32  traceD6.awk < D6_TRACE_FILE > OUTPUT_FILE

- Use Microsoft Excel to load the output file using '&' as delimiter.
- Sort by "Response Time" (Column B) to identify the queries with the longest response times.

4) trace_rpc_histD6.awk

This script is provided to help understand the response time contribution of groups of commands by analyzing a DFC trace file and summarizing (Total Time, Average Time, Total Calls) each RPC. It should be used to identify RPC's that take a short time for each individual call but they can be expensive because of repetitive usage.

- Execute the following command

    awk  trace_rpc_histD6.awk < D6_TRACE_FILE > OUTPUT_FILE
- The output file contains 3 sections: Summary Information, Profile of RPC Calls, Query + Response Time Information.