1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
|
#!/usr/bin/env expect
############################################################################
# Purpose: Test of Slurm functionality
# Test hdf5 acct_gather_profile (--profile=task)
############################################################################
# Copyright (C) 2013 Bull S. A. S.
# Bull, Rue Jean Jaures, B.P.68, 78340, Les Clayes-sous-Bois.
#
# Written by Rod Schultz <rod.schultz@bull.com>
#
# This file is part of Slurm, a resource management program.
# For details, see <https://slurm.schedmd.com/>.
# Please also read the included file: DISCLAIMER.
#
# Slurm is free software; you can redistribute it and/or modify it under
# the terms of the GNU General Public License as published by the Free
# Software Foundation; either version 2 of the License, or (at your option)
# any later version.
#
# Slurm is distributed in the hope that it will be useful, but WITHOUT ANY
# WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
# FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
# details.
#
# You should have received a copy of the GNU General Public License along
# with Slurm; if not, write to the Free Software Foundation, Inc.,
# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
############################################################################
source ./globals
set file_out "$test_dir/output"
set file_prog "$test_name.prog"
set hdf5_file "$test_name.h5"
set job_id 0
if {[get_config_param "FrontendName"] ne "MISSING"} {
skip "This test is incompatible with front-end systems"
}
proc cleanup {} {
global file_prog hdf5_file
file delete $file_prog $hdf5_file
}
# Check if acct_gather_profile/hdf5 is installed
set profile 0
log_user 0
spawn $scontrol show config
expect {
-re "acct_gather_profile/hdf5" {
set profile 1
}
eof {
wait
}
}
log_user 1
if {$profile == 0} {
skip "acct_gather_profile/hdf5 not installed on this system"
}
log_info "Acct_gather_profile/hdf5 plugin installed"
log_debug "Note: this test takes 3 minutes to run"
set task_freq [get_job_acct_freq]
if {$task_freq < 30} {
log_warn "jobacct_gather_freq < 30 ($task_freq), results are unreliable"
}
#
# Build a test program to put a known load on the system
#
exec $bin_rm -f $file_prog
exec $bin_cc -I$build_dir $file_prog.c -lm -o $file_prog
exec $bin_chmod 700 $file_prog
set timeout [expr $max_job_delay + 200]
spawn $srun --acctg-freq=$task_freq --profile=task -t5 ./$file_prog
expect {
-re "SLURM_JobId=($number)" {
set job_id $expect_out(1,string)
exp_continue
}
-re "error: PROFILE: Impossible to create the table" {
exec rm -f $file_prog
skip "This error is cause by a bug in HDF5 1.10.0, and is fixed in 1.10.1. Please upgrade to run this test"
}
-re "error:" {
fail "Something happened on start of job"
}
timeout {
fail "srun not responding"
}
eof {
wait
}
}
set timeout 10
spawn $sh5util -j $job_id
expect {
"Output file generated is empty" {
fail "sh5util merge didn't make anything"
}
timeout {
fail "sh5util merge not responding"
}
eof {
wait
}
}
spawn $sh5util -j $job_id -E -l Node:TimeSeries -s Tasks -o $file_out
expect {
"Output file generated is empty" {
fail "sh5util extract didn't make anything"
}
timeout {
fail "sh5util extract not responding"
}
eof {
wait
}
}
set line ""
set nerr 0
set lno 0
set fd 0
set last_et 0
set fd [open $file_out "r"]
set et_col -1
set cpu_util_col -1
set read_disk_col -1
while {$fd > 0 && [gets $fd line] != -1} {
incr lno
set tokens [split $line ","]
if {$lno == 1} {
set et_col [lsearch $tokens "ElapsedTime"]
set cpu_util_col [lsearch $tokens "CPUUtilization"]
set read_disk_col [lsearch $tokens "ReadMB"]
if {$et_col == -1} {
fail "No ElapsedTime column found"
}
if {$cpu_util_col == -1} {
fail "No CPUUtilization column found"
}
if {$read_disk_col == -1} {
fail "No ReadMB column found"
}
continue
}
set et [lindex $tokens $et_col]
set cur_et [expr $et - $last_et]
set last_et $et
if { $lno == 2 } {
continue
}
if {$cur_et < $task_freq} {
log_warn "Poll interval was only $cur_et instead of expected $task_freq on line $lno"
incr nerr
}
set cputil [lindex $tokens $cpu_util_col]
# The range on cpu utilization is pretty wide
# Linux accounting resolution is only to one second, so in a
# typical 30 interval an extra second is 3%. The burn loop
# consumes a bit more that asked for. There is additional type
# managing the I/O portion. Slurm and linux also consume some
# cpu.
if {$cputil < 38.0 || $cputil > 47.0 } {
log_warn "CPU Busy $cputil not near 40% on line $lno"
incr nerr
}
set rdmb [lindex $tokens $read_disk_col]
set low_rd [expr 0.975 * 10 * $cur_et]
set hi_rd [expr 1.025 * 10 * $cur_et]
if {$rdmb < $low_rd || $rdmb > $hi_rd } {
log_warn "Read Megabytes $rdmb not near 100.0 on line $lno"
incr nerr
}
}
close $fd
if {$lno == 0} {
fail "Output file ($file_out) is empty"
}
if {$nerr > 3} {
fail "Too many values out of range ($nerr)"
}
|
