commit 56fbd728c2df7f189390b9a417b4c3ecc3690ede Author: Nick Mathewson nickm@torproject.org Date: Tue May 3 16:45:15 2011 -0400

Backport microdesc_cache_clean to 0.2.2

Otherwise we have no way to keep authorities' microdesc caches in 0.2.2 from growing without bound. --- changes/bug2230_clean_1 | 4 +++ src/or/microdesc.c | 51 +++++++++++++++++++++++++++++++++++++++++++++++ src/or/microdesc.h | 1 + 3 files changed, 56 insertions(+), 0 deletions(-)

diff --git a/changes/bug2230_clean_1 b/changes/bug2230_clean_1 new file mode 100644 index 0000000..a4edf94 --- /dev/null +++ b/changes/bug2230_clean_1 @@ -0,0 +1,4 @@ + o Minor features + - Backport code from 0.2.3.x to allow directory authorities to clean + their microdescriptor caches. + diff --git a/src/or/microdesc.c b/src/or/microdesc.c index 0ceb134..58fbe37 100644 --- a/src/or/microdesc.c +++ b/src/or/microdesc.c @@ -23,6 +23,8 @@ struct microdesc_cache_t { tor_mmap_t *cache_content; /** Number of bytes used in the journal file. */ size_t journal_len; + /** Number of bytes in descriptors removed as too old. */ + size_t bytes_dropped;

/** Total bytes of microdescriptor bodies we have added to this cache */ uint64_t total_len_seen; @@ -276,6 +278,51 @@ microdesc_cache_reload(microdesc_cache_t *cache) return 0; }

+/** By default, we remove any microdescriptors that have gone at least this + * long without appearing in a current consensus. */ +#define TOLERATE_MICRODESC_AGE (7*24*60*60) + +/** Remove all microdescriptors from <b>cache</b> that haven't been listed for + * a long time. Does not rebuild the cache on disk. If <b>cutoff</b> is + * positive, specifically remove microdescriptors that have been unlisted + * since <b>cutoff</b>. If <b>force</b> is true, remove microdescriptors even + * if we have no current live microdescriptor consensus. + */ +void +microdesc_cache_clean(microdesc_cache_t *cache, time_t cutoff, int force) +{ + microdesc_t **mdp, *victim; + int dropped=0, kept=0; + size_t bytes_dropped = 0; + time_t now = time(NULL); + + (void) force; + /* In 0.2.2, we let this proceed unconditionally: only authorities have + * microdesc caches. */ + + if (cutoff <= 0) + cutoff = now - TOLERATE_MICRODESC_AGE; + + for (mdp = HT_START(microdesc_map, &cache->map); mdp != NULL; ) { + if ((*mdp)->last_listed < cutoff) { + ++dropped; + victim = *mdp; + mdp = HT_NEXT_RMV(microdesc_map, &cache->map, mdp); + bytes_dropped += victim->bodylen; + microdesc_free(victim); + } else { + ++kept; + mdp = HT_NEXT(microdesc_map, &cache->map, mdp); + } + } + + if (dropped) { + log_notice(LD_DIR, "Removed %d/%d microdescriptors as old.", + dropped,dropped+kept); + cache->bytes_dropped += bytes_dropped; + } +} + /** Regenerate the main cache file for <b>cache</b>, clear the journal file, * and update every microdesc_t in the cache with pointers to its new * location. */ @@ -291,6 +338,10 @@ microdesc_cache_rebuild(microdesc_cache_t *cache) int orig_size, new_size;

log_info(LD_DIR, "Rebuilding the microdescriptor cache..."); + + /* Remove dead descriptors */ + microdesc_cache_clean(cache, 0/*cutoff*/, 0/*force*/); + orig_size = (int)(cache->cache_content ? cache->cache_content->size : 0); orig_size += (int)cache->journal_len;

diff --git a/src/or/microdesc.h b/src/or/microdesc.h index b3e12f8..30cb25d 100644 --- a/src/or/microdesc.h +++ b/src/or/microdesc.h @@ -21,6 +21,7 @@ smartlist_t *microdescs_add_list_to_cache(microdesc_cache_t *cache, smartlist_t *descriptors, saved_location_t where, int no_save);

+void microdesc_cache_clean(microdesc_cache_t *cache, time_t cutoff, int force); int microdesc_cache_rebuild(microdesc_cache_t *cache); int microdesc_cache_reload(microdesc_cache_t *cache); void microdesc_cache_clear(microdesc_cache_t *cache);