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
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
|
# Read the source files and output the statistics #defines plus the
# initialize and refresh code.
import re, string, sys, textwrap
from dist import compare_srcfile
# Read the source files.
from stat_data import groups, dsrc_stats, connection_stats, join_stats
def print_struct(title, name, base, stats):
'''Print the structures for the stat.h file.'''
f.write('/*\n')
f.write(' * Statistics entries for ' + title + '.\n')
f.write(' */\n')
f.write('#define\tWT_' + name.upper() + '_STATS_BASE\t' + str(base) + '\n')
f.write('struct __wt_' + name + '_stats {\n')
for l in stats:
f.write('\tint64_t ' + l.name + ';\n')
f.write('};\n\n')
# Update the #defines in the stat.h file.
tmp_file = '__tmp'
f = open(tmp_file, 'w')
skip = 0
for line in open('../src/include/stat.h', 'r'):
if not skip:
f.write(line)
if line.count('Statistics section: END'):
f.write(line)
skip = 0
elif line.count('Statistics section: BEGIN'):
f.write('\n')
skip = 1
print_struct(
'connections', 'connection', 1000, connection_stats)
print_struct('data sources', 'dsrc', 2000, dsrc_stats)
print_struct('join cursors', 'join', 3000, join_stats)
f.close()
compare_srcfile(tmp_file, '../src/include/stat.h')
def print_defines_one(capname, base, stats):
for v, l in enumerate(stats, base):
f.write('/*! %s */\n' % '\n * '.join(textwrap.wrap(l.desc, 70)))
f.write('#define\tWT_STAT_' + capname + '_' + l.name.upper() + "\t" *
max(1, 6 - int((len('WT_STAT_' + capname + '_' + l.name)) / 8)) +
str(v) + '\n')
def print_defines():
'''Print the #defines for the wiredtiger.in file.'''
f.write('''
/*!
* @name Connection statistics
* @anchor statistics_keys
* @anchor statistics_conn
* Statistics are accessed through cursors with \c "statistics:" URIs.
* Individual statistics can be queried through the cursor using the following
* keys. See @ref data_statistics for more information.
* @{
*/
''')
print_defines_one('CONN', 1000, connection_stats)
f.write('''
/*!
* @}
* @name Statistics for data sources
* @anchor statistics_dsrc
* @{
*/
''')
print_defines_one('DSRC', 2000, dsrc_stats)
f.write('''
/*!
* @}
* @name Statistics for join cursors
* @anchor statistics_join
* @{
*/
''')
print_defines_one('JOIN', 3000, join_stats)
f.write('/*! @} */\n')
# Update the #defines in the wiredtiger.in file.
tmp_file = '__tmp'
f = open(tmp_file, 'w')
skip = 0
for line in open('../src/include/wiredtiger.in', 'r'):
if not skip:
f.write(line)
if line.count('Statistics section: END'):
f.write(line)
skip = 0
elif line.count('Statistics section: BEGIN'):
f.write(' */\n')
skip = 1
print_defines()
f.write('/*\n')
f.close()
compare_srcfile(tmp_file, '../src/include/wiredtiger.in')
def print_func(name, handle, list):
'''Print the structures/functions for the stat.c file.'''
f.write('\n')
f.write('static const char * const __stats_' + name + '_desc[] = {\n')
for l in list:
f.write('\t"' + l.desc + '",\n')
f.write('};\n')
f.write('''
int
__wt_stat_''' + name + '''_desc(WT_CURSOR_STAT *cst, int slot, const char **p)
{
\tWT_UNUSED(cst);
\t*p = __stats_''' + name + '''_desc[slot];
\treturn (0);
}
''')
f.write('''
void
__wt_stat_''' + name + '_init_single(WT_' + name.upper() + '''_STATS *stats)
{
\tmemset(stats, 0, sizeof(*stats));
}
''')
if handle != None:
f.write('''
void
__wt_stat_''' + name + '_init(' + handle + ''' *handle)
{
\tint i;
\tfor (i = 0; i < WT_COUNTER_SLOTS; ++i) {
\t\thandle->stats[i] = &handle->stat_array[i];
\t\t__wt_stat_''' + name + '''_init_single(handle->stats[i]);
\t}
}
''')
f.write('''
void
__wt_stat_''' + name + '_clear_single(WT_' + name.upper() + '''_STATS *stats)
{
''')
for l in sorted(list):
# no_clear: don't clear the value.
if 'no_clear' in l.flags:
f.write('\t\t/* not clearing ' + l.name + ' */\n')
else:
f.write('\tstats->' + l.name + ' = 0;\n')
f.write('}\n')
f.write('''
void
__wt_stat_''' + name + '_clear_all(WT_' + name.upper() + '''_STATS **stats)
{
\tu_int i;
\tfor (i = 0; i < WT_COUNTER_SLOTS; ++i)
\t\t__wt_stat_''' + name + '''_clear_single(stats[i]);
}
''')
# Single structure aggregation is currently only used by data sources.
if name == 'dsrc':
f.write('''
void
__wt_stat_''' + name + '''_aggregate_single(
WT_''' + name.upper() + '_STATS *from, WT_' + name.upper() + '''_STATS *to)
{
''')
for l in sorted(list):
if 'max_aggregate' in l.flags:
o = '\tif (from->' + l.name + ' > to->' + l.name + ')\n' +\
'\t\tto->' + l.name + ' = from->' + l.name + ';\n'
else:
o = '\tto->' + l.name + ' += from->' + l.name + ';\n'
if len(o) > 72: # Account for the leading tab.
o = o.replace(' += ', ' +=\n\t ')
f.write(o)
f.write('}\n')
f.write('''
void
__wt_stat_''' + name + '''_aggregate(
WT_''' + name.upper() + '_STATS **from, WT_' + name.upper() + '''_STATS *to)
{
''')
# Connection level aggregation does not currently have any computation
# of a maximum value; I'm leaving in support for it, but don't declare
# a temporary variable until it's needed.
for l in sorted(list):
if 'max_aggregate' in l.flags:
f.write('\tint64_t v;\n\n')
break;
for l in sorted(list):
if 'max_aggregate' in l.flags:
o = '\tif ((v = WT_STAT_READ(from, ' + l.name + ')) > ' +\
'to->' + l.name + ')\n'
if len(o) > 72: # Account for the leading tab.
o = o.replace(' > ', ' >\n\t ')
o +='\t\tto->' + l.name + ' = v;\n'
else:
o = '\tto->' + l.name + ' += WT_STAT_READ(from, ' + l.name + ');\n'
if len(o) > 72: # Account for the leading tab.
o = o.replace(' += ', ' +=\n\t ')
f.write(o)
f.write('}\n')
# Write the stat initialization and refresh routines to the stat.c file.
f = open(tmp_file, 'w')
f.write('/* DO NOT EDIT: automatically built by dist/stat.py. */\n\n')
f.write('#include "wt_internal.h"\n')
print_func('dsrc', 'WT_DATA_HANDLE', dsrc_stats)
print_func('connection', 'WT_CONNECTION_IMPL', connection_stats)
print_func('join', None, join_stats)
f.close()
compare_srcfile(tmp_file, '../src/support/stat.c')
# Update the statlog file with the entries we can scale per second.
scale_info = 'no_scale_per_second_list = [\n'
clear_info = 'no_clear_list = [\n'
prefix_list = []
for l in sorted(connection_stats):
prefix_list.append(l.prefix)
if 'no_scale' in l.flags:
scale_info += ' \'' + l.desc + '\',\n'
if 'no_clear' in l.flags:
clear_info += ' \'' + l.desc + '\',\n'
for l in sorted(dsrc_stats):
prefix_list.append(l.prefix)
if 'no_scale' in l.flags:
scale_info += ' \'' + l.desc + '\',\n'
if 'no_clear' in l.flags:
clear_info += ' \'' + l.desc + '\',\n'
# No join statistics can be captured in wtstats
scale_info += ']\n'
clear_info += ']\n'
prefix_info = 'prefix_list = [\n'
# Remove the duplicates and print out the list
for l in list(set(prefix_list)):
prefix_info += ' \'' + l + '\',\n'
prefix_info += ']\n'
group_info = 'groups = ' + str(groups)
tmp_file = '__tmp'
f = open(tmp_file, 'w')
f.write('# DO NOT EDIT: automatically built by dist/stat.py. */\n\n')
f.write(scale_info)
f.write(clear_info)
f.write(prefix_info)
f.write(group_info)
f.close()
compare_srcfile(tmp_file, '../tools/wtstats/stat_data.py')
|