config.py.sample 1.59 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
#! /usr/bin/env python3

# DEBUGGING
#############################
DEBUG = False

# INFLUXDB
#############################
# interval of measurements in seconds
METRIC_INTERVAL = 10

# names
measurements = {
    "proc": 'pfit-uprocstat',
    "jobs": 'pfit-jobinfo',
    "node": 'pfit-nodeinfo',
    "sys": 'system',
    "cpu": 'cpu',
    "infiniband": 'infiniband',
    "beegfs": 'beegfs_clients',
    "gpu_node": 'nvidia_gpu',
    "gpu_proc": 'nvidia_proc',
}

25
26
SLURM_BIN = "" # FULL path to Slurm binaries

27
28
29
30
31
32
33
34
35
# Security
#############################
LSF_BIN = ""
SECUSER = True # use setuid binary and check users before fetching data
job_uid_comm = {
    "lsf": "bjobs -noheader {jobid:s} -o \"user\" | id",
    "slurm": SLURM_BIN + "sacct -j {jobid:s} -X -P -n -o \"uid\"",
}

36
37
38
39
40
41
42
43
# Array JOBID transformation
#############################
JOBID_ARRAY_TRANSFORM = True
jobid_array_trans_comm = {
    "slurm": SLURM_BIN + "sacct -j {jobid:s} -X -P -n -o \"JobIDRaw\"",
    "lsf": "NOT SUPPORTED"
}

44
45
# DB
#############################
Azat Khuziyakhmetov's avatar
Azat Khuziyakhmetov committed
46
47
CACHEJSON = False # cache aggregated results as json file
CACHEDIR = "/path/to/caching/directory"
48
49
50

job_info = {
    "fetch_job_info_lsf": "bjobs -o \"{:s}\" {:s}",
51
    "fetch_job_info_slurm": SLURM_BIN + "sacct --format=\"{:s}\" -X -P -j {:s}",
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
    "measurement_name": "pfit-jobinfo",
}

# Batch system
BATCH_SYSTEM = "SLURM"  # LSF, SLURM
MIN_DUR = 50 # minimal duration of the job for generating report in minutes

# Network
INFINIBAND = True # aggregate infiniband data

# Filesystems
BEEGFS = True # aggregate beegfs data

# GPU
GPU = True # aggregat gpu data

# PDF report
SEQ_MAX_POINTS = 500