|
JVM Instances |
jvm_Backend_1(8)
|
OS Image Description |
os_1
|
Tuning |
- tuned-adm profile=throughput-performance
- echo 16000000 > /proc/sys/kernel/sched_latency_ns
- echo 200000 > /proc/sys/kernel/sched_min_granularity_ns
- echo 1000 > /proc/sys/kernel/sched_migration_cost_ns
- echo 3000000 > /proc/sys/kernel/sched_wakeup_granularity_ns
- Added cgroup_disable=memory,cpu,cpuacct,blkio,hugetlb,pids,cpuset,perf_event,freezer,devices,net_cls,net_prio to GRUB_CMDLINE_LINUX_DEFAULT
|
Notes |
None
|
Parts of Benchmark |
Backend
|
JVM Instance Description |
jvm_1
|
Command Line |
-XX:-UsePerfData -server -XX:AllocatePrefetchInstr=2 -XX:LargePageSizeInBytes=2m -XX:+AggressiveOpts -XX:-UseAdaptiveSizePolicy -XX:+AlwaysPreTouch -XX:-UseBiasedLocking -XX:+UseLargePages -XX:+UseParallelOldGC -Xms123g -Xmx123g -Xmn121g -XX:SurvivorRatio=80 -XX:TargetSurvivorRatio=48 -XX:ParallelGCThreads=16 -XX:MaxTenuringThreshold=15 -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:InlineSmallCode=12k -XX:MaxGCPauseMillis=500
|
Tuning |
None
|
Notes |
Used numactl to affinitize each Backend JVM to NUMA node. - numactl --cpunodebind=0 --localalloc
- numactl --cpunodebind=1 --localalloc
- numactl --cpunodebind=2 --localalloc
- numactl --cpunodebind=3 --localalloc
- numactl --cpunodebind=4 --localalloc
- numactl --cpunodebind=5 --localalloc
- numactl --cpunodebind=6 --localalloc
- numactl --cpunodebind=7 --localalloc
|
|