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
|
#! /bin/sh
# PCP QA Test No. 088
# interpolate mode, constant values, different logging rates
#
# Copyright (c) 1995-2002 Silicon Graphics, Inc. All Rights Reserved.
#
seq=`basename $0`
echo "QA output created by $seq"
# get standard filters
. ./common.product
. ./common.filter
_filter()
{
sed \
-e '/host:/d' \
-e '/start:/d' \
-e '/end:/d' \
-e 's/^..:..:..\..../TIMESTAMP /' \
| LC_COLLATE=POSIX sort \
| uniq
}
trap "rm -f $tmp.*; exit" 0 1 2 3 15
# real QA test starts here
for arch in tmparch/bar tmparch/mv-bar tmparch/noti-bar
do
pmval -U $arch \
-i bin-100,bin-500,bin-700,bin-900 sampledso.bin \
2>/dev/null | _filter
pmval -a $arch -t 0.5 \
-i bin-100,bin-500,bin-700,bin-900 sampledso.bin \
2>/dev/null | _filter
done
# Test for bug #500237
# (this should never print `end of log' - ie. there must be a timestamp,
# even though the single sample s1 can't have been rate converted).
# deal with an annoying libc error ... using TZ=:Australia/Melbourne
# causes the -z option to report the wrong time, at least for some
# times of the year
#
TZ=EST-10
export TZ
pmval -a archives/late -s1 -z sampledso.bin
|
