Slurm sacct/sacctmgr 命令

发布时间: 更新时间: 总字数:1373 阅读时间:3m 作者:IP:上海 网址
专栏文章
  1. Slurm 开发
  2. Slurm rpmbuild
  3. Slurm 相关术语
  4. Slurm QoS 使用
  5. Slurm Job
  6. Slurm Partition 使用
  7. Slurm REST API 使用
  8. Slurm scontrol 命令
  9. Slurm sacct/sacctmgr 命令(当前)
  10. Slurm 命令使用说明
  11. Slurm sreport 命令
  12. Slurm FaQ
  13. Slurm 安装部署

sacct/sacctmgr 在 Slurm 作业审计日志或 Slurm 数据库中显示所有作业、作业步骤的审计数据

ascct help

sacct--help ↕
sacct--helpformat ↕

sacct

查看已完成作业

bash
$ sacct
$ sacct -a
$ sacct -j <job-id>
$ sacct -j <job-id> -l  # 查看详细信息
$ sacct  -j 878 -l -p
JobID|JobIDRaw|JobName|Partition|MaxVMSize|MaxVMSizeNode|MaxVMSizeTask|AveVMSize|MaxRSS|MaxRSSNode|MaxRSSTask|AveRSS|MaxPages|MaxPagesNode|MaxPagesTask|AvePages|MinCPU|MinCPUNode|MinCPUTask|AveCPU|NTasks|AllocCPUS|Elapsed|State|ExitCode|AveCPUFreq|ReqCPUFreqMin|ReqCPUFreqMax|ReqCPUFreqGov|ReqMem|ConsumedEnergy|MaxDiskRead|MaxDiskReadNode|MaxDiskReadTask|AveDiskRead|MaxDiskWrite|MaxDiskWriteNode|MaxDiskWriteTask|AveDiskWrite|ReqTRES|AllocTRES|TRESUsageInAve|TRESUsageInMax|TRESUsageInMaxNode|TRESUsageInMaxTask|TRESUsageInMin|TRESUsageInMinNode|TRESUsageInMinTask|TRESUsageInTot|TRESUsageOutMax|TRESUsageOutMaxNode|TRESUsageOutMaxTask|TRESUsageOutAve|TRESUsageOutTot|
878|878|interactive|normal||||||||||||||||||2|00:01:15|FAILED|2:0||Unknown|Unknown|Unknown|2500Mc||||||||||billing=1,cpu=1,mem=2500M,node=1|billing=2,cpu=2,mem=5000M,node=1||||||||||||||

$ sacct -u <user-name>
$ sacct --starttime 2021-08-15
$ sacct --starttime 2021-08-15 --format=User,JobID,Jobname%50,partition,state,time,start,end,elapsed,MaxRss,MaxVMSize,nnodes,ncpus,nodelist

$ sacct -a --format  AllocCPUS,Partition,JobID
 AllocCPUS  Partition        JobID
---------- ---------- ------------
         2        poc 217

# 获取已经使用的CPU
$ sacct -s r --format Partition,NNodes,AllocCPUS
 Partition   NNodes  AllocCPUS
---------- -------- ----------
       poc        1          2

# 当前节点运行的作业
$ sacct -N <node-hostname> --format JobId --state running

# CPU 统计
sacct -j 115 -o CPUTimeRAW,CPUTime,TotalCPU --noheader
       148   00:02:28  01:12.284
    820710 9-11:58:30  00:01.416
       144   00:02:24  01:10.867

$ sstat -a -p -j <job_id> -o avecpu,avecpufreq,avediskread,avediskwrite

sacctmgr

  • View QOS, accounts, associations, TRES, wckeys, and users
  • Add/Remove/Update accounts, associations, and users
text
# 添加集群
sacctmgr add cluster myslurm

# 查看指定 user
[root@sms ~]$ sacctmgr show user root
      User   Def Acct     Admin
---------- ---------- ---------
      root       root      None
[root@sms ~]$ sacctmgr show user root -p
User|Def Acct|Admin|
root|root|None|
[root@sms ~]$ sacctmgr show user root -s -p
User|Def Acct|Admin|Cluster|Account|Partition|Share|Priority|MaxJobs|MaxNodes|MaxCPUs|MaxSubmit|MaxWall|MaxCPUMins|QOS|Def QOS|
root|root|None|poc|root||1||||||||root_qos||

# list user
$ sacctmgr list user -p

# list account
$ sacctmgr list account -s
$ sacctmgr list account -s -p
x

# 查看指定 account
[root@sms ~]$ sacctmgr show account root
   Account                Descr                  Org
---------- -------------------- --------------------
      root                 root                 root

[root@sms ~]$ sacctmgr show account root -s -p
Account|Descr|Org|Cluster|Par Name|User|Share|Priority|GrpJobs|GrpNodes|GrpCPUs|GrpMem|GrpSubmit|GrpWall|GrpCPUMins|MaxJobs|MaxNodes|MaxCPUs|MaxSubmit|MaxWall|MaxCPUMins|QOS
|Def QOS|
root|root|root|poc|root||1|||||||||||||||normal||
root|root|root|poc||user1|1|||||||||||||||normal||
root|root|root|poc||root|1|||||||||||||||root_qos||

# 查看 qos
[root@sms ~]$ sacctmgr show qos normal -s -p
Name|Priority|GraceTime|Preempt|PreemptExemptTime|PreemptMode|Flags|UsageThres|UsageFactor|GrpTRES|GrpTRESMins|GrpTRESRunMins|GrpJobs|GrpSubmit|GrpWall|MaxTRES|MaxTRESPerNode|MaxTRESMins|MaxWall|MaxTRESPU|MaxJobsPU|MaxSubmitPU|MaxTRESPA|MaxJobsPA|MaxSubmitPA|MinTRES|
normal|0|00:00:00|||cluster|||1.000000||||||||||||||||||

# 查询账户
sacctmgr show account
sacctmgr show account withassoc

# 添加账户
sacctmgr add account user1
[ Parent=<root> ]

# 修改账户
sacctmgr modify account user1 set GrpJobs=2
sacctmgr modify account user1 set Parent=root

# 删除账号
sacctmgr delete account user1

# 查看 tres
$ sacctmgr list tres
    Type            Name     ID
-------- --------------- ------
     cpu                      1
     mem                      2
  energy                      3
    node                      4
 billing                      5
      fs            disk      6
    vmem                      7
   pages                      8
    gres             gpu   1004

create user demo

text
# 添加 QoS 并修改对应信息
sacctmgr add qos user2_qos
sacctmgr modify qos user2_qos set MaxJobsPerUser=4
sacctmgr modify qos user2_qos set MaxTRESPerUser=CPU=2
sacctmgr modify qos user2_qos set MaxJobsPU=2
sacctmgr modify qos user2_qos set MaxSubmitJobsPerUser=2

# 添加 account
sacctmgr add account user2 Description="user2" Organization=user2

# 添加 user
sacctmgr add user user2 Account=user2 QOS=user2_qos Cluster=poc

# 删除 user
sacctmgr remove user user2