SPECjbb2015
Copyright © 2015-2017 Standard Performance Evaluation Corporation
H3C H3C UniServer R4900 G3 175588 SPECjbb2015-MultiJVM max-jOPS
64011 SPECjbb2015-MultiJVM critical-jOPS
Tested by: H3C Test Sponsor: H3C Test location: Hangzhou, ZheJiang, China Test date: Aug 17, 2017
SPEC license #: 9066 Hardware Availability: Jul-2017 Software Availability: Jul-2017 Publication: Thu Sep 28 17:00:08 EDT 2017
Benchmark Results Summary
 
SPECjbb2015-MultiJVM: Multiple JVMs/Single Host
(# of groups: 4)
Overall Throughput RT curve
Overall Throughput RT curve
Overall SUT (System Under Test) Description
VendorH3C
Vendor URLhttp://www.h3c.com
System SourceSingle Supplier
System DesignationServer Rack
Total Systems1
All SUT Systems IdenticalYES
Total Nodes1
All Nodes IdenticalYES
Nodes Per System1
Total Chips2
Total Cores56
Total Threads112
Total Memory Amount (GB)384
Total OS Images1
SW EnvironmentNon-virtual
  
SUT Description
Hardware hw_1
NameH3C UniServer R4900 G3
VendorH3C
Vendor URLhttp://www.h3c.com
AvailableJul-2017
ModelH3C UniServer R4900 G3
Form Factor2U
CPU NameIntel(R) Xeon(R) Platinum 8180M CPU
CPU Characteristics28-Core, 2.50GHz, 38.5MB L3 Cache (Turbo Boost Technology up to 3.80 GHz)
Number of Systems1
Nodes Per System1
Chips Per System2
Cores Per System56
Cores Per Chip28
Threads Per System112
Threads Per Core2
VersionNone
CPU Frequency (MHz)2500
Primary Cache32 KB(I)+32 KB(D) per core
Secondary Cache1024 KB (I+D) per core
Tertiary Cache39424 KB (I+D) on chip per chip
Other CacheNone
Disk1 x 480 GB SSD SATA
File SystemXFS
Memory Amount (GB)384
# and size of DIMM(s)24 x 16GB
Memory Details384 GB (24 x 16 GB 2Rx8 PC4-2666V-R)
# and type of Network Interface Cards (NICs)2 x 10 Gbit NICs
Power Supply Quantity and Rating (W)2 x 500
Other HardwareNone
Cabinet/Housing/EnclosureNone
Shared DescriptionNone
Shared CommentNone
NotesNone
Other Hardware network_1
NameNone
VendorNone
Vendor URLNone
VersionNone
AvailableNone
Bitnessn/a
NotesNone
Operating System os_1
NameSUSE Linux Enterprise Server 12 SP2
VendorSUSE
Vendor URLhttp://www.suse.com
VersionSUSE Linux Enterprise Server 12 SP2 (x86_64)
AvailableNov-2016
Bitness64
NotesNone
Java Virtual Machine jvm_1
NameOracle Java SE 8u144
VendorOracle Corporation
Vendor URLhttp://www.oracle.com
VersionJava HotSpot 64-Bit Server VM, version 1.8.0_144
AvailableJul-2017
Bitness64
NotesNone
Other Software other_1
NameNone
VendorNone
Vendor URLNone
VersionNone
AvailableNone
BitnessNone
NotesNone
Topology
SUT config_1 Configuration
Hardware
OS Images os_Image_1(1)
Hardware Description hw_1
Number of Systems 1
SW Environment non-virtual
Tuning BIOS tuning
  • SNC was enabled
  • IMC set to 1-way Interleave
  • Energy Performance BIAS Mode set to Performance
Notes None
OS Image os_Image_1
JVM Instances jvm_Ctr_1(1), jvm_Backend_1(4), jvm_TxInjector_1(4)
OS Image Description os_1
Tuning
  • echo 990000 > /proc/sys/kernel/sched_rt_runtime_us
  • echo 1000 > /proc/sys/kernel/sched_migration_cost_ns
  • echo 16000000 > /proc/sys/kernel/sched_latency_ns
  • echo 6000000 > /proc/sys/kernel/sched_min_granularity_ns
Notes None
JVM Instance jvm_Ctr_1
Parts of Benchmark Controller
JVM Instance Description jvm_1
Command Line -Xms2g -Xmx2g -Xmn1536m -XX:+UseParallelOldGC
Tuning None
Notes None
JVM Instance jvm_Backend_1
Parts of Benchmark Backend
JVM Instance Description jvm_1
Command Line -showversion -Xms30g -Xmx30g -Xmn28g -server -XX:AllocatePrefetchInstr=2 -XX:LargePageSizeInBytes=2m -XX:-UsePerfData -XX:+AggressiveOpts -XX:-UseAdaptiveSizePolicy -XX:+AlwaysPreTouch -XX:-UseBiasedLocking -XX:+UseLargePages -XX:+UseParallelOldGC -XX:SurvivorRatio=28 -XX:TargetSurvivorRatio=95 -XX:ParallelGCThreads=28 -XX:MaxTenuringThreshold=15 -Xnoclassgc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+UseRTMLocking -XX:InlineSmallCode=10k -XX:+UseRTMDeopt -XX:MaxGCPauseMillis=300
Tuning numactl was used to affinitize each Backend JVM to a single numa node, eg:
  • numactl --cpunodebind=$node --localalloc
Notes None
JVM Instance jvm_TxInjector_1
Parts of Benchmark TxInjector
JVM Instance Description jvm_1
Command Line -Xms2g -Xmx2g -Xmn1536m -XX:+UseParallelOldGC
Tuning numactl was used to affinitize each TxInjector JVM to a single numa node, eg:
  • numactl --cpunodebind=$node --localalloc
Notes None
max-jOPS and critical-jOPS Details
max-jOPS = jOPS passed before the First Failure
Pass/Fail Pass Pass Fail Fail Fail
jOPS 173615 175588 177561 179534 181507
critical-jOPS = Geomean ( jOPS @ 10000; 25000; 50000; 75000; 100000; SLAs )
Response time percentile is 99-th
SLA (us) 10000 25000 50000 75000 100000 Geomean
jOPS 44390 50637 64119 77929 95685 64011
Last Success jOPS/First Failure jOPS for SLA points
  Percentile
  10-th 50-th 90-th 95-th 99-th 100-th
500us 1973 / 3946 - / 1973 - / 1973 - / 1973 - / 1973 - / 1973
1000us 23675 / 25648 5919 / 7892 - / 1973 - / 1973 - / 1973 - / 1973
5000us 161778 / 163751 147967 / 149940 88780 / 86808 47350 / 49322 29593 / 31566 1973 / 3946
10000us 163751 / 165724 151913 / 153886 118374 / 120347 92726 / 94699 43404 / 45377 5919 / 3946
25000us 165724 / 167696 157832 / 159805 134157 / 136130 118374 / 116401 53268 / 49322 5919 / 3946
50000us 165724 / 167696 159805 / 161778 140076 / 142049 122320 / 124293 63133 / 65106 5919 / 3946
75000us 165724 / 167696 159805 / 161778 142049 / 144022 126266 / 128238 78916 / 76943 5919 / 3946
100000us 169669 / 171642 161778 / 163751 145995 / 147967 132184 / 134157 96672 / 94699 5919 / 3946
200000us 171642 / 173615 165724 / 167696 155859 / 157832 151913 / 153886 140076 / 142049 72997 / 17756
500000us 175588 / - 171642 / 173615 163751 / 165724 163751 / 165724 159805 / 161778 145995 / 147967
1000000us 175588 / - 175588 / - 171642 / 173615 171642 / 173615 169669 / 171642 161778 / 163751
Number of probes
Probes jOPS / Total jOPS
Request Mix Accuracy
Request Mix Accuracy
Note
(Actual % in the Mix - Expected % in the Mix) must be within:
'Main Tx' limit of +/-5.0% for the requests whose expected % in the mix is >= 10.0%
'Minor Tx' limit of +/-1.0% for the requests whose expected % in the mix is < 10.0%
Rate of Non-Critical Failures
There were no non-critical failures in Response Time curve building
Delay between performance status pings during RT Curve (Response-Throughput Curve)
Delay between status pings
IR/PR Accuracy
IR/PR Accuracy
Run Properties
This section lists properties only set by user
Property Name Default Controller Group1.Backend.beJVM Group1.TxInjector.txiJVM1 Group2.Backend.beJVM Group2.TxInjector.txiJVM1 Group3.Backend.beJVM Group3.TxInjector.txiJVM1 Group4.Backend.beJVM Group4.TxInjector.txiJVM1
specjbb.comm.connect.client.pool.size 256 192
specjbb.comm.connect.selector.runner.count 0 2
specjbb.comm.connect.timeouts.connect 60000 700000 600000 600000 600000 600000 600000 600000 600000 600000
specjbb.comm.connect.timeouts.read 60000 700000 600000 600000 600000 600000 600000 600000 600000 600000
specjbb.comm.connect.timeouts.write 60000 700000 600000 600000 600000 600000 600000 600000 600000 600000
specjbb.comm.connect.worker.pool.max 256 64
specjbb.comm.connect.worker.pool.min 1 2
specjbb.controller.maxir.maxFailedPoints 3 3
specjbb.controller.rtcurve.warmup.step 0.1 0.1
specjbb.customerDriver.threads 64 {probe=64, saturate=64}
specjbb.forkjoin.workers 112 {=112, Tier1=200, Tier2=75, Tier3=32}
specjbb.group.count 1 4
specjbb.heartbeat.threshold 100000 200000
specjbb.mapreducer.pool.size 112 112
specjbb.txi.pergroup.count 1 1
View table in csv format
Validation Details
Validation Reports
 
Level: COMPLIANCE
Check Agent Result
Check properties on compliance All PASSED
 
Level: CORRECTNESS
Check Agent Result
Compare SM and HQ Inventory All PASSED
Other Checks
High-bound (max attempted) is 197290 IR
High-bound (settled) is 185592 IR
Copyright © 2015-2017 Standard Performance Evaluation Corporation
http://www.spec.org - info@spec.org
SPECjbb2015 Version: [SPECjbb2015 1.00, June 24, 2015]