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
|
#!/bin/sh
# PCP QA Test No. 1794
# Exercise pcp2arrow(1) parquet file generation.
#
# Copyright (c) 2024 Red Hat. All Rights Reserved.
#
seq=`basename $0`
echo "QA output created by $seq"
. ./common.python
test -x $PCP_BIN_DIR/pcp2arrow || _notrun "pcp2arrow(1) is not installed"
$python -c "import pandas" >/dev/null 2>&1 || _notrun "pandas is not installed"
_cleanup()
{
cd $here
$sudo rm -rf $tmp $tmp.*
}
status=0 # success is the default!
trap "_cleanup; exit \$status" 0 1 2 3 15
_filter()
{
sed \
-e "s@$tmp@TMP@g" \
-e "s@0.0[0-9][0-9]*@SMALLNUM@g" \
-e "s@0.0@ZERONUM@g" \
# end
}
# real QA test starts here
for archive in viewqa1 viewqa2 viewqa3 moomba.client proc
do
echo
echo === pcp2arrow $archive
pcp2arrow -z -t 10 -o $tmp.$archive -a archives/$archive
find $tmp.$archive >> $seq_full
$python -c "
import pandas
df = pandas.read_parquet('$tmp.$archive')
if (len(df.columns) < 42):
print('Columns:', sorted(df.columns))
# exercise bug in extracting instance values
if 'kernel.all.load[1 minute]' in df.columns:
print('1 and 15 minute load averages')
ldf = df[['kernel.all.load[1 minute]', 'kernel.all.load[15 minute]']]
print(ldf.head())
else:
print('Columns:', len(df.columns))
print('Shape:', df.shape)
" | _filter
done
# Test raw mode (pmFetchArchive optimization)
echo
echo "=== Testing raw mode with --raw flag ==="
for archive in viewqa1 moomba.client
do
echo
echo === pcp2arrow --raw $archive
pcp2arrow -z -r -s 5 -o $tmp.raw.$archive -a archives/$archive kernel.all.load disk.all.read
find $tmp.raw.$archive >> $seq_full
$python -c "
import pandas
df = pandas.read_parquet('$tmp.raw.$archive')
print('Columns:', sorted(df.columns))
print('Shape:', df.shape)
print('Has timestamp:', 'timestamp' in df.columns)
# Verify timestamps are valid (not 1970 epoch)
if 'timestamp' in df.columns and len(df) > 0:
ts = df['timestamp'].iloc[0]
if ts.year >= 1999:
print('Timestamp valid: True')
else:
print('Timestamp valid: False (year=%d)' % ts.year)
" | _filter
done
# success, all done
exit
|
