[or-cvs] r15345: implement more fine-tuning options for stats code (tor/trunk/src/or)

[nickm@xxxxxxxx]

Author: nickm
Date: 2008-06-18 00:34:52 -0400 (Wed, 18 Jun 2008)
New Revision: 15345

Modified:
   tor/trunk/src/or/config.c
   tor/trunk/src/or/geoip.c
   tor/trunk/src/or/or.h
   tor/trunk/src/or/router.c
Log:
implement more fine-tuning options for stats code

Modified: tor/trunk/src/or/config.c
===================================================================
--- tor/trunk/src/or/config.c	2008-06-18 04:30:00 UTC (rev 15344)
+++ tor/trunk/src/or/config.c	2008-06-18 04:34:52 UTC (rev 15345)
@@ -183,6 +183,9 @@
   OBSOLETE("DirPostPeriod"),
 #ifdef ENABLE_GEOIP_STATS
   V(DirRecordUsageByCountry,     BOOL,     "0"),
+  V(DirRecordUsageGranularity,   UINT,     "4"),
+  V(DirRecordUsageRetainIPs,     INTERVAL, "14 days"),
+  V(DirRecordUsageSaveInterval,  INTERVAL, "6 hours"),
 #endif
   VAR("DirServer",               LINELIST, DirServers, NULL),
   V(DNSPort,                     UINT,     "0"),

Modified: tor/trunk/src/or/geoip.c
===================================================================
--- tor/trunk/src/or/geoip.c	2008-06-18 04:30:00 UTC (rev 15344)
+++ tor/trunk/src/or/geoip.c	2008-06-18 04:34:52 UTC (rev 15345)
@@ -408,6 +408,14 @@
 /*DOCDOC*/
 #define GEOIP_MIN_OBSERVATION_TIME (12*60*60)
 
+static INLINE unsigned
+round_to_next_multiple_of(unsigned number, unsigned divisor)
+{
+  number += divisor - 1;
+  number -= number % divisor;
+  return number;
+}
+
 /** Return a newly allocated comma-separated string containing entries for all
  * the countries from which we've seen enough clients connect. The entry
  * format is cc=num where num is the number of IPs we've seen connecting from
@@ -428,6 +436,11 @@
     clientmap_entry_t **ent;
     unsigned *counts = tor_malloc_zero(sizeof(unsigned)*n_countries);
     unsigned total = 0;
+    unsigned granularity = IP_GRANULARITY;
+#ifdef ENABLE_GEOIP_STATS
+    if (get_options()->DirRecordUsageByCountry)
+      granularity = get_options()->DirRecordUsageGranularity;
+#endif
     HT_FOREACH(ent, clientmap, &client_history) {
       int country;
       if (((*ent)->last_seen & ACTION_MASK) != action)
@@ -456,9 +469,7 @@
 #else
       if (c > 0) {
 #endif
-        /* Round up to the next multiple of IP_GRANULARITY */
-        c += IP_GRANULARITY-1;
-        c -= c % IP_GRANULARITY;
+        c = round_to_next_multiple_of(c, granularity);
         countrycode = geoip_get_country_name(i);
         ent = tor_malloc(sizeof(c_hist_t));
         strlcpy(ent->country, countrycode, sizeof(ent->country));
@@ -493,12 +504,18 @@
   return result;
 }
 
-  /**DOCDOC*/
+/**DOCDOC*/
 char *
 geoip_get_request_history(time_t now, geoip_client_action_t action)
 {
-  smartlist_t *entries;
+  smartlist_t *entries, *strings;
   char *result;
+  unsigned granularity = IP_GRANULARITY;
+#ifdef ENABLE_GEOIP_STATS
+  if (get_options()->DirRecordUsageByCountry)
+    granularity = get_options()->DirRecordUsageGranularity;
+#endif
+
   if (client_history_starts >= (now - GEOIP_MIN_OBSERVATION_TIME))
     return NULL;
   if (action != GEOIP_CLIENT_NETWORKSTATUS &&
@@ -506,23 +523,36 @@
     return NULL;
   if (!geoip_countries)
     return NULL;
+
   entries = smartlist_create();
   SMARTLIST_FOREACH(geoip_countries, geoip_country_t *, c, {
       uint32_t *n = (action == GEOIP_CLIENT_NETWORKSTATUS)
         ? c->n_v3_ns_requests : c->n_v2_ns_requests;
       uint32_t tot = 0;
       int i;
-      char buf[32];
+      c_hist_t *ent;
       for (i=0; i < REQUEST_HIST_LEN; ++i)
         tot += n[i];
       if (!tot)
         continue;
-      tor_snprintf(buf, sizeof(buf), "%s=%ld", c->countrycode, (long)tot);
-      smartlist_add(entries, tor_strdup(buf));
+      ent = tor_malloc_zero(sizeof(c_hist_t));
+      strlcpy(ent->country, c->countrycode, sizeof(ent->country));
+      ent->total = round_to_next_multiple_of(tot, granularity);
+      smartlist_add(entries, ent);
   });
-  smartlist_sort_strings(entries);
-  result = smartlist_join_strings(entries, ",", 0, NULL);
-  SMARTLIST_FOREACH(entries, char *, cp, tor_free(cp));
+  smartlist_sort(entries, _c_hist_compare);
+
+  strings = smartlist_create();
+  SMARTLIST_FOREACH(entries, c_hist_t *, ent, {
+      char buf[32];
+      tor_snprintf(buf, sizeof(buf), "%s=%u", ent->country, ent->total);
+      smartlist_add(strings, tor_strdup(buf));
+    });
+  result = smartlist_join_strings(strings, ",", 0, NULL);
+  SMARTLIST_FOREACH(strings, char *, cp, tor_free(cp));
+  SMARTLIST_FOREACH(entries, c_hist_t *, ent, tor_free(ent));
+  smartlist_free(strings);
+  smartlist_free(entries);
   return result;
 }
 

Modified: tor/trunk/src/or/or.h
===================================================================
--- tor/trunk/src/or/or.h	2008-06-18 04:30:00 UTC (rev 15344)
+++ tor/trunk/src/or/or.h	2008-06-18 04:34:52 UTC (rev 15345)
@@ -2396,6 +2396,9 @@
   int BridgeRecordUsageByCountry;
 #ifdef ENABLE_GEOIP_STATS
   int DirRecordUsageByCountry;
+  int DirRecordUsageGranularity;
+  int DirRecordUsageRetainIPs;
+  int DirRecordUsageSaveInterval;
 #endif
 
   /** Optionally, a file with GeoIP data. */

Modified: tor/trunk/src/or/router.c
===================================================================
--- tor/trunk/src/or/router.c	2008-06-18 04:30:00 UTC (rev 15344)
+++ tor/trunk/src/or/router.c	2008-06-18 04:34:52 UTC (rev 15345)
@@ -1826,8 +1826,13 @@
     static time_t last_purged_at = 0;
     char *geoip_summary;
     time_t now = time(NULL);
-    if (now > last_purged_at+48*60*60) {
-      geoip_remove_old_clients(now-48*60*60);
+    int geoip_purge_interval = 48*60*60;
+#ifdef ENABLE_GEOIP_STATS
+    if (get_options()->DirRecordUsageByCountry)
+      geoip_purge_interval = get_options()->DirRecordUsageRetainIPs;
+#endif
+    if (now > last_purged_at+geoip_purge_interval) {
+      geoip_remove_old_clients(now-geoip_purge_interval);
       last_purged_at = now;
     }
     geoip_summary = geoip_get_client_history(time(NULL), GEOIP_CLIENT_CONNECT);