Add site configs and harden portal handling

Signed-off-by: Yoshifumi Nakamura <nakamura@riken.jp>

Author: yoshifuminakamura

.github/workflows/gitlab-manual-ci.yml

@@ -44,19 +44,27 @@ jobs:
     name: Run GitLab CI manually
     runs-on: ubuntu-latest
     steps:
-      - name: Check out target ref
+      - name: Check out trusted workflow ref
         uses: actions/checkout@v4
         with:
           fetch-depth: 0
-          ref: ${{ inputs.target_ref }}
+          path: trusted
 
       - name: Prepare GitLab repository settings
         id: gitlab-repo
-        uses: ./.github/actions/prepare-gitlab-repo
+        uses: ./trusted/.github/actions/prepare-gitlab-repo
         with:
           gitlab-repo: ${{ secrets.GITLAB_REPO }}
 
+      - name: Check out target ref
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+          ref: ${{ inputs.target_ref }}
+          path: target
+
       - name: Push target ref to GitLab test branch
+        working-directory: target
         env:
           GITLAB_TOKEN: ${{ secrets.GITLAB_TOKEN }}
           GITLAB_REPO_HOST_PATH: ${{ steps.gitlab-repo.outputs.host-path }}
@@ -184,6 +192,7 @@ jobs:
 
       - name: Delete GitLab test branch
         if: always()
+        working-directory: trusted
         env:
           GITLAB_TOKEN: ${{ secrets.GITLAB_TOKEN }}
           GITLAB_REPO_HOST_PATH: ${{ steps.gitlab-repo.outputs.host-path }}

benchpark-bridge/scripts/result_converter.py

@@ -10,7 +10,6 @@
 import os
 import sys
 import glob
-import yaml
 from datetime import datetime
 from pathlib import Path
 
@@ -535,4 +534,4 @@ def main():
 
 
 if __name__ == "__main__":
-    main()
+    main()

config/queue.csv

@@ -1,4 +1,5 @@
 queue,submit_cmd,template
+SLURM_AI4SS,sbatch,"-p ${queue_group} -t ${elapse} -N ${nodes} --ntasks-per-node=${numproc_node} --cpus-per-task=${nthreads} --gpus-per-node=${numproc_node}"
 FJ,pjsub,"-L rscunit=rscunit_ft01,rscgrp=${queue_group},elapse=${elapse},node=${nodes} --mpi max-proc-per-node=${numproc_node} -x PJM_LLIO_GFSCACHE=/vol0002:/vol0003:/vol0004:/vol0005"
 PJM_GENKAI,pjsub,"-L rscgrp=${queue_group},elapse=${elapse},node=${nodes} --mpi proc=${proc}"
 SLURM_RC,sbatch,"-p ${queue_group} -t ${elapse} -N ${nodes} --ntasks-per-node=${numproc_node} --cpus-per-task=${nthreads}"
@@ -7,4 +8,12 @@ PBS_Grand_C,qsub,"-q ${queue_group} -l select=${nodes}:nsockets=${cpu_per_node},
 PBS_Grand_G,qsub,"-q ${queue_group} -l select=${nodes}:ngpus=1,walltime=${elapse} -W group_list=d30992"
 NQSV_AOBA_VE,qsub,"-Z -v http_proxy,https_proxy,HTTP_PROXY,HTTPS_PROXY -q ${queue_group} -T necmpi --venode ${proc} -l elapstim_req=${elapse}"
 NQSV_AOBA_B,qsub,"-Z -v http_proxy,https_proxy,HTTP_PROXY,HTTPS_PROXY -q ${queue_group} -T intmpi -b ${nodes} -l elapstim_req=${elapse}"
+PJM_WISTERIA_O,pjsub,"-g jh260034o -L rscgrp=${queue_group},elapse=${elapse},node=${nodes} --mpi proc=${proc} --omp thread=${nthreads}"
+PJM_WISTERIA_A,pjsub,"-g jh260034a -L rscgrp=${queue_group},elapse=${elapse},node=${nodes} --mpi proc=${proc} --omp thread=${nthreads}"
+PBS_TSUKUBA,qsub,"-q ${queue_group} -l select=${nodes}:mpiprocs=${numproc_node}:ompthreads=${nthreads} -l walltime=${elapse}"
+AGE_TSUBAME4,qsub,"-l ${queue_group}=${nodes} -l h_rt=${elapse}"
+SLURM_CAMPHOR3,sbatch,"-p ${queue_group} -t ${elapse} --rsc p=${proc}:t=${nthreads}:c=${nthreads}:m=1G"
+NQSV_OSAKA_CPU,qsub,"-q ${queue_group} -b ${nodes} -l elapstim_req=${elapse},cpunum_job=${nthreads}"
+NQSV_OSAKA_GPU,qsub,"-q ${queue_group} -b ${nodes} -l elapstim_req=${elapse},cpunum_job=${nthreads},gpunum_job=${gpu_per_node}"
+NQSV_OSAKA_VE,qsub,"-q ${queue_group} --venode ${proc} -l elapstim_req=${elapse}"
 none,none,none

config/system.csv

@@ -1,4 +1,5 @@
 system,mode,tag_build,tag_run,queue,queue_group
+AI4SS,cross,ai4ss_login,ai4ss_jacamar,SLURM_AI4SS,1n1gpu
 Fugaku,cross,fugaku_login1,fugaku_jacamar,FJ,small
 FugakuLN,native,,fugaku_login1,none,small
 FugakuCN,native,,fugaku_jacamar,FJ,small
@@ -16,4 +17,14 @@ Grand_G,cross,grand_login,grand_jacamar,PBS_Grand_G,eg
 AOBA_A,cross,aoba_ab_login,aoba_ab_jacamar,NQSV_AOBA_VE,sx
 AOBA_B,cross,aoba_ab_login,aoba_ab_jacamar,NQSV_AOBA_B,lx
 AOBA_S,cross,aoba_s_login,aoba_s_jacamar,NQSV_AOBA_VE,sxs
+Odyssey,cross,wisteria_login,wisteria-o_jacamar,PJM_WISTERIA_O,short-o
+Aquarius,cross,wisteria_login,wisteria-a_jacamar,PJM_WISTERIA_A,short-a
+Pegasus,cross,pegasus_login,pegasus_jacamar,PBS_TSUKUBA,regular
+Sirius,cross,sirius_login,sirius_jacamar,PBS_TSUKUBA,regular
+TSUBAME4,cross,tsubame4_login,tsubame4_jacamar,AGE_TSUBAME4,node_f
+Camphor3,cross,camphor3_login,camphor3_jacamar,SLURM_CAMPHOR3,jha
+SQUID_CPU,cross,squid_login,squid_jacamar,NQSV_OSAKA_CPU,SQUID
+SQUID_GPU,cross,squid_login,squid_jacamar,NQSV_OSAKA_GPU,SQUID
+SQUID_VECTOR,cross,squid_login,squid_jacamar,NQSV_OSAKA_VE,SQUID
+OCTOPUS,cross,octopus_login,octopus_jacamar,NQSV_OSAKA_CPU,OCT
 FNCX,native,,fncx-curl-jq,none,small

config/system_info.csv

@@ -1,18 +1,29 @@
 system,name,cpu_name,cpu_per_node,cpu_cores,gpu_name,gpu_per_node,memory,display_order
-Fugaku,Fugaku,A64FX,1,48,-,-,32GB,1
-FugakuCN,FugakuCN,A64FX,1,48,-,-,32GB,2
-FugakuLN,FugakuLN,Intel(R) Xeon(R) Gold 6242 CPU @ 2.80GHz,2,16,-,-,96GB,3
-MiyabiG,MiyabiG,NVIDIA Grace CPU,1,72,NVIDIA Hopper H100 GPU,1,120GB,4
-MiyabiC,MiyabiC,Intel Xeon Max 9480,2,56,-,-,128GB,5
-RC_GH200,RC_GH200,NVIDIA Grace CPU,1,72,NVIDIA Hopper H100 GPU,1,120GB,6
-RC_DGXSP,RC_DGXSP,ARM Cortex-X925 / Cortex-A725,1,20,NVIDIA GB10,1,128GB,7
-RC_GENOA,RC_GENOA,AMD EPYC 9684X,2,96,-,-,768GB,8
-RC_FX700,RC_FX700,A64FX,1,48,-,-,32GB,9
-GenkaiA,GenkaiA,Intel Xeon Platinum 8490H (Sapphire Rapids),2,60,-,-,512GiB,10
-GenkaiB,GenkaiB,Intel Xeon Platinum 8490H (Sapphire Rapids),2,60,NVIDIA H100 (Hopper),4,1024GiB,11
-GenkaiC,GenkaiC,Intel Xeon Platinum 8480+ (Sapphire Rapids),2,56,NVIDIA H100 (Hopper),8,8TiB,12
-Grand_C,Grand_C,Intel Xeon Gold 6548Y+ (Emerald Rapids),2,32,-,-,512GiB,13
-Grand_G,Grand_G,Intel Xeon Gold 6548Y+ (Emerald Rapids),2,32,NVIDIA H100 (Hopper),4,512GiB,14
-AOBA_A,AOBA_A,SX-Aurora TSUBASA VH,1,24,NEC SX-Aurora TSUBASA Type 20B VE,8,640GB,15
-AOBA_B,AOBA_B,AMD EPYC 7702,2,64,-,-,256GB,16
-AOBA_S,AOBA_S,SX-Aurora TSUBASA VH,1,64,NEC SX-Aurora TSUBASA Type 30A VE,8,256GB + 768GB,17
+AI4SS,RIKEN AI4S Supercomputer,NVIDIA Grace CPU,2,72,NVIDIA B200,4,960GiB + 692.8GiB,1
+Fugaku,Fugaku,A64FX,1,48,-,-,32GB,2
+FugakuCN,FugakuCN,A64FX,1,48,-,-,32GB,3
+FugakuLN,FugakuLN,Intel(R) Xeon(R) Gold 6242 CPU @ 2.80GHz,2,16,-,-,96GB,4
+MiyabiG,MiyabiG,NVIDIA Grace CPU,1,72,NVIDIA Hopper H100 GPU,1,120GB,5
+MiyabiC,MiyabiC,Intel Xeon Max 9480,2,56,-,-,128GB,6
+RC_GH200,RC_GH200,NVIDIA Grace CPU,1,72,NVIDIA Hopper H100 GPU,1,120GB,7
+RC_DGXSP,RC_DGXSP,ARM Cortex-X925 / Cortex-A725,1,20,NVIDIA GB10,1,128GB,8
+RC_GENOA,RC_GENOA,AMD EPYC 9684X,2,96,-,-,768GB,9
+RC_FX700,RC_FX700,A64FX,1,48,-,-,32GB,10
+GenkaiA,GenkaiA,Intel Xeon Platinum 8490H (Sapphire Rapids),2,60,-,-,512GiB,11
+GenkaiB,GenkaiB,Intel Xeon Platinum 8490H (Sapphire Rapids),2,60,NVIDIA H100 (Hopper),4,1024GiB,12
+GenkaiC,GenkaiC,Intel Xeon Platinum 8480+ (Sapphire Rapids),2,56,NVIDIA H100 (Hopper),8,8TiB,13
+Grand_C,Grand_C,Intel Xeon Gold 6548Y+ (Emerald Rapids),2,32,-,-,512GiB,14
+Grand_G,Grand_G,Intel Xeon Gold 6548Y+ (Emerald Rapids),2,32,NVIDIA H100 (Hopper),4,512GiB,15
+AOBA_A,AOBA_A,SX-Aurora TSUBASA VH,1,24,NEC SX-Aurora TSUBASA Type 20B VE,8,640GB,16
+AOBA_B,AOBA_B,AMD EPYC 7702,2,64,-,-,256GB,17
+AOBA_S,AOBA_S,SX-Aurora TSUBASA VH,1,64,NEC SX-Aurora TSUBASA Type 30A VE,8,256GB + 768GB,18
+Odyssey,Odyssey,A64FX,1,48,-,-,32GiB,19
+Aquarius,Aquarius,Intel Xeon Platinum 8360Y,2,36,NVIDIA A100,8,512GiB,20
+TSUBAME4,TSUBAME4.0,AMD EPYC 9654,2,96,NVIDIA H100 SXM5 94GB HBM2e,4,768GiB,21
+Camphor3,Camphor3,Intel Xeon CPU Max 9480,2,56,-,-,128GiB,22
+Pegasus,Pegasus,Intel Xeon Platinum 8468,2,48,NVIDIA H100 PCIe,1,128GiB + 2TiB PMem,23
+Sirius,Sirius (PACS12.0),AMD EPYC Zen 4 (MI300A APU),4,24,AMD Instinct MI300A CDNA3,4,512GB HBM3,24
+SQUID_CPU,SQUID CPU,Intel Xeon Platinum 8368,2,38,-,-,256GB,25
+SQUID_GPU,SQUID GPU,Intel Xeon Platinum 8368,2,38,NVIDIA A100 SXM4 40GB,8,512GB,26
+SQUID_VECTOR,SQUID Vector,AMD EPYC 7402P,1,24,NEC SX-Aurora TSUBASA Type20A VE,8,128GB + 384GB,27
+OCTOPUS,OCTOPUS,Intel Xeon 6980P (Granite Rapids),2,128,-,-,768GB,28

docs/ci.md

@@ -260,7 +260,7 @@ Use these examples when deciding whether to split a pull request or start GitLab
 | Example change set / 変更例 | Expected checks / 期待される確認 | GitLab benchmark expectation / GitLab benchmark期待値 |
 |---|---|---|
 | `docs/ci.md` only / `docs/ci.md`のみ | Review the documentation diff / docs差分をreview | No benchmark run. Direct/manual GitLab pipelines should skip by rules / benchmark不要。直接/手動GitLab pipelineではrulesでskipされる想定 |
-| `result_server/routes/usage.py` and `result_server/templates/*.html` / `result_server/routes/usage.py`と`result_server/templates/*.html` | `Result Server Tests` should run / `Result Server Tests`が動く | No benchmark run unless a maintainer intentionally starts one / maintainerが意図して起動しない限りbenchmark不要 |
+| `result_server/routes/results_usage_routes.py` and `result_server/templates/*.html` / `result_server/routes/results_usage_routes.py`と`result_server/templates/*.html` | `Result Server Tests` should run / `Result Server Tests`が動く | No benchmark run unless a maintainer intentionally starts one / maintainerが意図して起動しない限りbenchmark不要 |
 | `config/system_info.csv` only / `config/system_info.csv`のみ | `Result Server Tests` should verify public site config consistency / 公開site config整合性を`Result Server Tests`で確認 | No benchmark run because this file is portal display metadata / portal表示metadataなのでbenchmark不要 |
 | `config/system.csv` or `config/queue.csv` for a public system / 公開system向けの`config/system.csv`または`config/queue.csv` | `Result Server Tests` should run the site config preflight / `Result Server Tests`でsite config preflightを実行 | Start `GitLab Manual CI` too when benchmark execution behavior needs validation / benchmark実行挙動の検証が必要なら`GitLab Manual CI`も起動 |
 | `scripts/bk_functions.sh`, `scripts/result.sh`, or `scripts/result_server/**` only / `scripts/bk_functions.sh`、`scripts/result.sh`、または`scripts/result_server/**`のみ | `Result Server Tests` should run when the path filter matches / path filter対象なら`Result Server Tests`が動く | Manual GitLab CI is optional and only needed if upload behavior affects benchmark operation / upload挙動がbenchmark運用に影響する場合だけ手動GitLab CIを検討 |

programs/qws/build.sh

@@ -36,6 +36,10 @@ case "$system" in
 	echo "Dummy build for FNCX Docker runner test"
 	echo aaaa > main
 	;;
+    AI4SS)
+	module load nvhpc-hpcx/26.3
+	make -j 8 omp=1 compiler=nvhpc-hpcx arch=grace rdma= mpi=1
+	;;
     RC_GH200)
 	module load system/qc-gh200 nvhpc-hpcx/25.9
 	### QWSはNeoverse版やGPU版はないので汎用版としてとりあえずarch=skylakeを指定している
@@ -75,6 +79,38 @@ case "$system" in
     AOBA_B)
 	make -j 8 fugaku_benchmark= omp=1 compiler=openmpi-gnu arch=skylake rdma= mpi=1 powerapi= CC=mpicc CXX=mpic++
 	;;
+    Odyssey)
+	module load odyssey
+	make compiler=fujitsu_cross arch=postk -j 8
+	;;
+    Aquarius)
+	module purge
+	module load intel
+	source /work/opt/local/x86_64/cores/intel/2023.0.0/mpi/latest/env/vars.sh
+	make compiler=intel arch=skylake rdma= -j8
+	;;
+    TSUBAME4)
+	make -j 8 fugaku_benchmark= omp=1 compiler=openmpi-gnu arch=skylake rdma= mpi=1 powerapi= CC=mpicc CXX=mpic++
+	;;
+    Camphor3)
+	camphor3_modulepath="${MODULEPATH:-}"
+	if [[ -r /etc/profile.d/modules.sh ]]; then
+	    source /etc/profile.d/modules.sh
+	elif [[ -r /etc/profile.d/z00_lmod.sh ]]; then
+	    source /etc/profile.d/z00_lmod.sh
+	else
+	    echo "qws: no module init script found" >&2
+	fi
+	if [[ -n "${MODULEPATH:-}" ]]; then
+	    camphor3_modulepath="${MODULEPATH}"
+	fi
+	module purge
+	if [[ -n "${camphor3_modulepath:-}" ]]; then
+	    export MODULEPATH="${camphor3_modulepath}"
+	fi
+	module load slurm/2022 SysA/2022 intel/2023.2 intelmpi/2023.2 PrgEnvIntel/2023
+	make -j 8 fugaku_benchmark= omp=1 compiler=intel arch=skylake rdma= mpi=1 powerapi=
+	;;
     *)
 	echo "Unknown system: $system"
 	exit 1

programs/qws/list.csv

@@ -1,12 +1,13 @@
 system,enable,nodes,numproc_node,nthreads,elapse
+AI4SS,yes,1,1,72,0:10:00
 Fugaku,yes,1,4,12,0:10:00
 FugakuLN,yes,1,1,1,0:10:00
 FugakuCN,no,1,4,12,0:10:00
 FugakuCN,no,2,4,12,0:10:00
 RC_GH200,yes,1,1,72,0:10:00
 RC_DGXSP,yes,1,1,20,0:10:00
 RC_GENOA,yes,1,1,96,0:10:00
-RC_FX700,yes,1,4,12,0:10:00
+RC_FX700,yes,1,1,12,0:10:00
 MiyabiG,yes,1,1,72,0:10:00
 MiyabiC,yes,1,1,112,0:10:00
 GenkaiA,yes,1,1,120,0:10:00
@@ -17,4 +18,8 @@ Grand_G,yes,1,1,64,0:10:00
 AOBA_A,yes,1,1,8,0:10:00
 AOBA_S,yes,1,1,8,0:10:00
 AOBA_B,yes,1,1,128,0:10:00
+Odyssey,yes,1,1,12,0:10:00
+Aquarius,yes,1,1,8,0:10:00
+TSUBAME4,yes,1,1,192,0:10:00
+Camphor3,yes,1,1,112,0:10:00
 FNCX,yes,1,1,1,0:10:00

programs/qws/run.sh

@@ -84,6 +84,14 @@ case "$system" in
         echo 'dummy call for FNCX Docker runner test'
         bk_emit_result --fom 99.99 --fom-version dummy --exp FNCXTest --nodes "$nodes" --numproc-node "$numproc_node" --nthreads "$nthreads" >> ../results/result
         ;;
+    AI4SS)
+        module load nvhpc-hpcx/26.3
+        export OMP_NUM_THREADS=72
+        export OMP_PLACES=cores
+        export OMP_PROC_BIND=close
+        mpirun --bind-to none -n 1 ./main 32 6 4 3 1 1 1 1 -1 -1 6 50 > CASE0
+        print_results CASE0 CASE0 1 >> ../results/result
+        ;;
     RC_GH200)
         module load system/qc-gh200 nvhpc-hpcx/25.9
         mpirun -n 1 --bind-to core --map-by ppr:1:node:PE=72 ./main 32 6 4 3 1 1 1 1 -1 -1 6 50 > CASE0
@@ -141,6 +149,58 @@ case "$system" in
         mpirun -np ${qws_numproc} ./main 32 6 4 3 1 1 1 1 -1 -1 6 50 > CASE0
         print_results CASE0 CASE0 ${numproc_node} >> ../results/result
         ;;
+    Odyssey)
+        if [[ -r /etc/profile.d/modules.sh ]]; then
+            source /etc/profile.d/modules.sh
+        else
+            echo "qws: /etc/profile.d/modules.sh is not readable" >&2
+        fi
+        module unload fjmpi fj odyssey 2>/dev/null || true
+        module load odyssey fj fjmpi
+        export OMP_NUM_THREADS=12
+        export PLE_MPI_STD_EMPTYFILE=off
+        mpiexec -n 1 -ofout CASE0 ./main 32 6 4 3 1 1 1 1 -1 -1 6 50
+        print_results CASE0 CASE0 1 >> ../results/result
+        ;;
+    Aquarius)
+        module purge
+        module load intel
+        source /work/opt/local/x86_64/cores/intel/2023.0.0/mpi/latest/env/vars.sh
+        export OMP_NUM_THREADS=8
+        export I_MPI_PIN=1
+        mpiexec -n 1 ./main 32 6 4 3 1 1 1 1 -1 -1 6 50 > CASE0
+        print_results CASE0 CASE0 1 >> ../results/result
+        ;;
+    TSUBAME4)
+        qws_numproc=$((nodes * numproc_node))
+        mpirun -n ${qws_numproc} ./main 32 6 4 3 1 1 1 1 -1 -1 6 50 > CASE0
+        print_results CASE0 CASE0 ${numproc_node} >> ../results/result
+        ;;
+    Camphor3)
+        camphor3_modulepath="${MODULEPATH:-}"
+        if [[ -r /etc/profile.d/modules.sh ]]; then
+            source /etc/profile.d/modules.sh
+        elif [[ -r /etc/profile.d/z00_lmod.sh ]]; then
+            source /etc/profile.d/z00_lmod.sh
+        else
+            echo "qws: no module init script found" >&2
+        fi
+        if [[ -n "${MODULEPATH:-}" ]]; then
+            camphor3_modulepath="${MODULEPATH}"
+        fi
+        module purge
+        if [[ -n "${camphor3_modulepath:-}" ]]; then
+            export MODULEPATH="${camphor3_modulepath}"
+        fi
+        module load intel/2023.2 intelmpi/2023.2 PrgEnvIntel/2023
+        export OMP_NUM_THREADS="${nthreads}"
+        export I_MPI_PIN=1
+        if [[ "${SLURM_CONF:-}" == /etc/slurm/sysA/* ]]; then
+            unset SLURM_CONF
+        fi
+        srun -n 1 -c "${nthreads}" ./main 32 6 4 3 1 1 1 1 -1 -1 6 50 > CASE0
+        print_results CASE0 CASE0 1 >> ../results/result
+        ;;
     *)
         echo "Unknown Running system: $system"
         exit 1

requirements-result-server.txt

@@ -1,4 +1,4 @@
-# result_server requires Python 3.12+ for safe tar extraction via tarfile filter="data".
+# result_server is tested and deployed on Python 3.12+; archive handling uses explicit path/type validation.
 Flask>=3.0,<4.0
 Flask-Session>=0.8,<1.0
 Flask-WTF>=1.2,<2.0