|
| JVM Instances |
jvm_Ctr_1(1), jvm_Backend_1(4), jvm_TxInjector_1(4)
|
| OS Image Description |
os_1
|
| Tuning |
A processor set was configured containing the virtual processors (HW threads) of CPU 1 (command: "psrset -c 12-23 36-47").
|
| Notes |
The controller and groups 3 and 4 (backend and transaction injector) were run in the processor set (command: "psrset -e 1 priocntl -e -c FX java ..."). Groups 1 and 2 (backend and transaction injector) were run outside the processor set (command: "priocntl -e -c FX java ..."). The "priocntl -e -c FX" was used in these commands to run them with the FX scheduling class.
|
|
| Parts of Benchmark |
Controller
|
| JVM Instance Description |
jvm_1
|
| Command Line |
-Xmx2048m -Xms2048m -Xmn1536m -Dspecjbb.forkjoin.workers=24 -Dspecjbb.group.count=2
|
| Tuning |
A processor set was configured containing the virtual processors (HW threads) of CPU 1.
|
| Notes |
Controller was run in processor set 1 with FX scheduling class. The object-caching memory allocation library libumem was used by setting the environment variable LD_PRELOAD=libumem.so.
|
|
| Parts of Benchmark |
Backend
|
| JVM Instance Description |
jvm_1
|
| Command Line |
-Xmx22g -Xms22g -Xmn17g -XX:+UseParallelOldGC -XX:ParallelGCThreads=14 -XX:+UseCompressedOops -XX:-UseBiasedLocking -XX:+UseLargePages -XX:LargePageSizeInBytes=2m
|
| Tuning |
A processor set was configured containing the virtual processors (HW threads) of CPU 1.
|
| Notes |
The controller and groups 3 and 4 (backend and transaction injector) were run in the processor set. Groups 1 and 2 (backend and transaction injector) were run outside the processor set, with the FX scheduling class. The object-caching memory allocation library libumem was used by setting the environment variable LD_PRELOAD=libumem.so.
|
|
| Parts of Benchmark |
TxInjector
|
| JVM Instance Description |
jvm_1
|
| Command Line |
-Xmx2048m -Xms2048m -Xmn1536m
|
| Tuning |
A processor set was configured containing the virtual processors (HW threads) of CPU 1.
|
| Notes |
The controller and groups 3 and 4 (backend and transaction injector) were run in the processor set. Groups 1 and 2 (backend and transaction injector) were run outside the processor set, with the FX scheduling class. The object-caching memory allocation library libumem was used by setting the environment variable LD_PRELOAD=libumem.so.
|
