]>
Commit | Line | Data |
---|---|---|
f740a279 | 1 | |
8667b856 | 2 | /* |
c3031d67 | 3 | * $Id: cache_diff.cc,v 1.21 2007/11/15 16:47:35 wessels Exp $ |
8667b856 | 4 | * |
5 | * AUTHOR: Alex Rousskov | |
6 | * | |
2b6662ba | 7 | * SQUID Web Proxy Cache http://www.squid-cache.org/ |
e25c139f | 8 | * ---------------------------------------------------------- |
8667b856 | 9 | * |
2b6662ba | 10 | * Squid is the result of efforts by numerous individuals from |
11 | * the Internet community; see the CONTRIBUTORS file for full | |
12 | * details. Many organizations have provided support for Squid's | |
13 | * development; see the SPONSORS file for full details. Squid is | |
14 | * Copyrighted (C) 2001 by the Regents of the University of | |
15 | * California; see the COPYRIGHT file for full details. Squid | |
16 | * incorporates software developed and/or copyrighted by other | |
17 | * sources; see the CREDITS file for full details. | |
8667b856 | 18 | * |
19 | * This program is free software; you can redistribute it and/or modify | |
20 | * it under the terms of the GNU General Public License as published by | |
21 | * the Free Software Foundation; either version 2 of the License, or | |
22 | * (at your option) any later version. | |
23 | * | |
24 | * This program is distributed in the hope that it will be useful, | |
25 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
26 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
27 | * GNU General Public License for more details. | |
28 | * | |
29 | * You should have received a copy of the GNU General Public License | |
30 | * along with this program; if not, write to the Free Software | |
cbdec147 | 31 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA. |
e25c139f | 32 | * |
8667b856 | 33 | */ |
34 | ||
35 | /* | |
36 | * Computes the difference between the contents of two caches | |
37 | * using swap logs | |
38 | * Reports the percentage of common files and other stats | |
39 | */ | |
40 | ||
41 | #include "squid.h" | |
42 | ||
62e76326 | 43 | typedef struct |
44 | { | |
8667b856 | 45 | const char *name; |
46 | hash_table *hash; | |
1afe05c5 | 47 | int count; /* #currently cached entries */ |
48 | int scanned_count; /* #scanned entries */ | |
49 | int bad_add_count; /* #duplicate adds */ | |
50 | int bad_del_count; /* #dels with no prior add */ | |
62e76326 | 51 | } |
52 | ||
53 | CacheIndex; | |
8667b856 | 54 | |
55 | ||
62e76326 | 56 | typedef struct _CacheEntry |
57 | { | |
8667b856 | 58 | const cache_key *key; |
62e76326 | 59 | |
8667b856 | 60 | struct _CacheEntry *next; |
51ee7c82 | 61 | /* StoreSwapLogData s; */ |
c3031d67 | 62 | unsigned char key_arr[SQUID_MD5_DIGEST_LENGTH]; |
62e76326 | 63 | } |
64 | ||
65 | CacheEntry; | |
8667b856 | 66 | |
67 | ||
68 | /* copied from url.c */ | |
69 | const char *RequestMethodStr[] = | |
62e76326 | 70 | { |
71 | "NONE", | |
72 | "GET", | |
73 | "POST", | |
74 | "PUT", | |
75 | "HEAD", | |
76 | "CONNECT", | |
77 | "TRACE", | |
78 | "PURGE" | |
79 | }; | |
8667b856 | 80 | |
81 | ||
1afe05c5 | 82 | static int cacheIndexScan(CacheIndex * idx, const char *fname, FILE * file); |
8667b856 | 83 | |
84 | ||
85 | static CacheEntry * | |
51ee7c82 | 86 | cacheEntryCreate(const StoreSwapLogData * s) |
8667b856 | 87 | { |
88 | CacheEntry *e = xcalloc(1, sizeof(CacheEntry)); | |
89 | assert(s); | |
edc1d6da | 90 | /* e->s = *s; */ |
c3031d67 | 91 | xmemcpy(e->key_arr, s->key, SQUID_MD5_DIGEST_LENGTH); |
edc1d6da | 92 | e->key = &e->key_arr[0]; |
8667b856 | 93 | return e; |
94 | } | |
95 | ||
96 | static void | |
1afe05c5 | 97 | cacheEntryDestroy(CacheEntry * e) |
8667b856 | 98 | { |
99 | assert(e); | |
100 | xfree(e); | |
101 | } | |
102 | ||
103 | static CacheIndex * | |
104 | cacheIndexCreate(const char *name) | |
105 | { | |
106 | CacheIndex *idx; | |
62e76326 | 107 | |
8667b856 | 108 | if (!name || !strlen(name)) |
62e76326 | 109 | return NULL; |
8667b856 | 110 | |
111 | idx = xcalloc(1, sizeof(CacheIndex)); | |
62e76326 | 112 | |
8667b856 | 113 | idx->name = name; |
62e76326 | 114 | |
8667b856 | 115 | idx->hash = hash_create(storeKeyHashCmp, 2e6, storeKeyHashHash); |
116 | ||
117 | return idx; | |
118 | } | |
119 | ||
120 | static void | |
1afe05c5 | 121 | cacheIndexDestroy(CacheIndex * idx) |
8667b856 | 122 | { |
123 | hash_link *hashr = NULL; | |
62e76326 | 124 | |
8667b856 | 125 | if (idx) { |
62e76326 | 126 | /* destroy hash list contents */ |
127 | hash_first(idx->hash); | |
128 | ||
129 | while (hashr = hash_next(idx->hash)) { | |
130 | hash_remove_link(idx->hash, hashr); | |
131 | cacheEntryDestroy((CacheEntry *) hashr); | |
132 | } | |
133 | ||
134 | /* destroy the hash table itself */ | |
135 | hashFreeMemory(idx->hash); | |
136 | ||
137 | xfree(idx); | |
8667b856 | 138 | } |
139 | } | |
140 | ||
141 | static int | |
1afe05c5 | 142 | cacheIndexAddLog(CacheIndex * idx, const char *fname) |
8667b856 | 143 | { |
144 | FILE *file; | |
145 | int scanned_count = 0; | |
146 | assert(idx); | |
147 | assert(fname && strlen(fname)); | |
148 | ||
149 | file = fopen(fname, "r"); | |
62e76326 | 150 | |
8667b856 | 151 | if (!file) { |
62e76326 | 152 | fprintf(stderr, "cannot open %s: %s\n", fname, strerror(errno)); |
153 | return 0; | |
8667b856 | 154 | } |
62e76326 | 155 | |
ec4daaa5 | 156 | #ifdef _SQUID_WIN32_ |
c4aefe96 | 157 | setmode(fileno(file), O_BINARY); |
62e76326 | 158 | |
c4aefe96 | 159 | #endif |
62e76326 | 160 | |
8667b856 | 161 | scanned_count = cacheIndexScan(idx, fname, file); |
162 | ||
163 | fclose(file); | |
62e76326 | 164 | |
8667b856 | 165 | return scanned_count; |
166 | } | |
167 | ||
168 | static void | |
1afe05c5 | 169 | cacheIndexInitReport(CacheIndex * idx) |
8667b856 | 170 | { |
171 | assert(idx); | |
172 | fprintf(stderr, "%s: bad swap_add: %d\n", | |
62e76326 | 173 | idx->name, idx->bad_add_count); |
1afe05c5 | 174 | fprintf(stderr, "%s: bad swap_del: %d\n", |
62e76326 | 175 | idx->name, idx->bad_del_count); |
1afe05c5 | 176 | fprintf(stderr, "%s: scanned lines: %d\n", |
62e76326 | 177 | idx->name, idx->scanned_count); |
8667b856 | 178 | } |
179 | ||
180 | static int | |
1afe05c5 | 181 | cacheIndexScan(CacheIndex * idx, const char *fname, FILE * file) |
8667b856 | 182 | { |
183 | int count = 0; | |
51ee7c82 | 184 | StoreSwapLogData s; |
8667b856 | 185 | fprintf(stderr, "%s scanning\n", fname); |
62e76326 | 186 | |
8667b856 | 187 | while (fread(&s, sizeof(s), 1, file) == 1) { |
62e76326 | 188 | count++; |
189 | idx->scanned_count++; | |
190 | /* if (s.op <= SWAP_LOG_NOP || s.op >= SWAP_LOG_MAX) | |
191 | * continue; */ | |
192 | ||
193 | if (s.op == SWAP_LOG_ADD) { | |
194 | CacheEntry *olde = (CacheEntry *) hash_lookup(idx->hash, s.key); | |
195 | ||
196 | if (olde) { | |
197 | idx->bad_add_count++; | |
198 | } else { | |
199 | CacheEntry *e = cacheEntryCreate(&s); | |
200 | hash_join(idx->hash, &e->hash); | |
201 | idx->count++; | |
202 | } | |
203 | } else if (s.op == SWAP_LOG_DEL) { | |
204 | CacheEntry *olde = (CacheEntry *) hash_lookup(idx->hash, s.key); | |
205 | ||
206 | if (!olde) | |
207 | idx->bad_del_count++; | |
208 | else { | |
209 | assert(idx->count); | |
210 | hash_remove_link(idx->hash, (hash_link *) olde); | |
211 | cacheEntryDestroy(olde); | |
212 | idx->count--; | |
213 | } | |
214 | } else { | |
215 | fprintf(stderr, "%s:%d: unknown swap log action\n", fname, count); | |
216 | exit(-3); | |
217 | } | |
8667b856 | 218 | } |
62e76326 | 219 | |
1afe05c5 | 220 | fprintf(stderr, "%s:%d: scanned (size: %d bytes)\n", |
62e76326 | 221 | fname, count, (int) (count * sizeof(CacheEntry))); |
8667b856 | 222 | return count; |
223 | } | |
224 | ||
225 | static void | |
1afe05c5 | 226 | cacheIndexCmpReport(CacheIndex * idx, int shared_count) |
8667b856 | 227 | { |
228 | assert(idx && shared_count <= idx->count); | |
229 | ||
f97c9706 | 230 | printf("%s:\t %7d = %7d + %7d (%7.2f%% + %7.2f%%)\n", |
62e76326 | 231 | idx->name, |
232 | idx->count, | |
233 | idx->count - shared_count, | |
234 | shared_count, | |
235 | xpercent(idx->count - shared_count, idx->count), | |
236 | xpercent(shared_count, idx->count)); | |
8667b856 | 237 | } |
238 | ||
239 | static void | |
1afe05c5 | 240 | cacheIndexCmp(CacheIndex * idx1, CacheIndex * idx2) |
8667b856 | 241 | { |
242 | int shared_count = 0; | |
681e7924 | 243 | int hashed_count = 0; |
8667b856 | 244 | hash_link *hashr = NULL; |
245 | CacheIndex *small_idx = idx1; | |
246 | CacheIndex *large_idx = idx2; | |
247 | assert(idx1 && idx2); | |
248 | ||
249 | /* check our guess */ | |
62e76326 | 250 | |
8667b856 | 251 | if (idx1->count > idx2->count) { |
62e76326 | 252 | small_idx = idx2; |
253 | large_idx = idx1; | |
8667b856 | 254 | } |
62e76326 | 255 | |
8667b856 | 256 | /* find shared_count */ |
0f6bebac | 257 | hash_first(small_idx->hash); |
62e76326 | 258 | |
0f6bebac | 259 | for (hashr = hash_next(small_idx->hash)) { |
62e76326 | 260 | hashed_count++; |
261 | ||
262 | if (hash_lookup(large_idx->hash, hashr->key)) | |
263 | shared_count++; | |
8667b856 | 264 | } |
62e76326 | 265 | |
681e7924 | 266 | assert(hashed_count == small_idx->count); |
8667b856 | 267 | |
268 | cacheIndexCmpReport(idx1, shared_count); | |
269 | cacheIndexCmpReport(idx2, shared_count); | |
270 | } | |
271 | ||
272 | ||
273 | static int | |
274 | usage(const char *prg_name) | |
275 | { | |
276 | fprintf(stderr, "usage: %s <label1>: <swap_state>... <label2>: <swap_state>...\n", | |
62e76326 | 277 | prg_name); |
8667b856 | 278 | return -1; |
279 | } | |
280 | ||
281 | int | |
282 | main(int argc, char *argv[]) | |
283 | { | |
284 | CacheIndex *CacheIdx[2]; | |
285 | CacheIndex *idx = NULL; | |
286 | int idxCount = 0; | |
287 | int i; | |
288 | ||
289 | if (argc < 5) | |
62e76326 | 290 | return usage(argv[0]); |
8667b856 | 291 | |
292 | for (i = 1; i < argc; ++i) { | |
62e76326 | 293 | const int len = strlen(argv[i]); |
294 | ||
295 | if (!len) | |
296 | return usage(argv[0]); | |
297 | ||
298 | if (argv[i][len - 1] == ':') { | |
299 | idxCount++; | |
300 | ||
301 | if (len < 2 || idxCount > 2) | |
302 | return usage(argv[0]); | |
303 | ||
304 | idx = cacheIndexCreate(argv[i]); | |
305 | ||
306 | CacheIdx[idxCount - 1] = idx; | |
307 | } else { | |
308 | if (!idx) | |
309 | return usage(argv[0]); | |
310 | ||
311 | cacheIndexAddLog(idx, argv[i]); | |
312 | } | |
8667b856 | 313 | } |
314 | ||
315 | if (idxCount != 2) | |
62e76326 | 316 | return usage(argv[0]); |
8667b856 | 317 | |
318 | cacheIndexInitReport(CacheIdx[0]); | |
62e76326 | 319 | |
8667b856 | 320 | cacheIndexInitReport(CacheIdx[1]); |
321 | ||
322 | cacheIndexCmp(CacheIdx[0], CacheIdx[1]); | |
323 | ||
324 | cacheIndexDestroy(CacheIdx[0]); | |
62e76326 | 325 | |
8667b856 | 326 | cacheIndexDestroy(CacheIdx[1]); |
327 | ||
328 | return 1; | |
329 | } |