|
|
|
| JVM Instances |
jvm_Backend_1(8)
|
| OS Image Description |
os_1
|
| Tuning |
- tuned-adm profile=throughput-performance
- echo 10000 > /proc/sys/vm/dirty_expire_centisecs
- echo 1500 > /proc/sys/vm/dirty_writeback_centisecs
- echo 990000 > /proc/sys/kernel/sched_rt_runtime_us
- echo always > /sys/kernel/mm/transparent_hugepage/defrag
- echo 1000 > /proc/sys/kernel/sched_migration_cost_ns
- systemctl stop systemd-update-utmp-runlevel.service
- 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 |
-showversion -server -XX:AllocatePrefetchInstr=2 -XX:LargePageSizeInBytes=2m -XX:-UsePerfData -XX:+AggressiveOpts -XX:-UseAdaptiveSizePolicy -XX:+AlwaysPreTouch -XX:-UseBiasedLocking -XX:+UseLargePages -XX:+UseParallelOldGC -Xms30g -Xmx30g -Xmn28g -XX:SurvivorRatio=28 -XX:TargetSurvivorRatio=98 -XX:ParallelGCThreads=16 -XX:MaxTenuringThreshold=15 -Xnoclassgc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:InlineSmallCode=10k -XX:MaxGCPauseMillis=300 -XX:ThreadStackSize=1m
|
| Tuning |
None
|
| Notes |
Used numactl to affinitize each Backend JVM to physical cores in a NUMA node. - numactl --physcpubind=0-7,64-71 --membind=0
- numactl --physcpubind=8-15,72-79 --membind=1
- numactl --physcpubind=16-23,80-87 --membind=2
- numactl --physcpubind=24-31,88-95 --membind=3
- numactl --physcpubind=32-39,96-103 --membind=4
- numactl --physcpubind=40-47,104-111 --membind=5
- numactl --physcpubind=48-55,112-119 --membind=6
- numactl --physcpubind=56-63,120-127 --membind=7
|
|
