[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
[or-cvs] [tor/release-0.2.2 022/162] handle ugly edge case in retrying entrynodes
Author: Roger Dingledine <arma@xxxxxxxxxxxxxx>
Date: Tue, 28 Sep 2010 22:32:38 -0400
Subject: handle ugly edge case in retrying entrynodes
Commit: 9997676802c140aceddb849090c7b3795fc83361
Specifically, a circ attempt that we'd launched while the network was
down could timeout after we've marked our entrynodes up, marking them
back down again. The fix is to annotate as bad the OR conns that were
around before we did the retry, so if a circuit that's attached to them
times out we don't do anything about it.
---
src/or/circuitbuild.c | 11 ++++++++++-
src/or/circuituse.c | 9 ++++++++-
src/or/connection_or.c | 16 ++++++++++------
src/or/connection_or.h | 2 +-
src/or/main.c | 2 +-
5 files changed, 30 insertions(+), 10 deletions(-)
diff --git a/src/or/circuitbuild.c b/src/or/circuitbuild.c
index a9920e2..dd92e78 100644
--- a/src/or/circuitbuild.c
+++ b/src/or/circuitbuild.c
@@ -4604,7 +4604,16 @@ entries_retry_helper(or_options_t *options, int act)
any_known = 1;
if (ri->is_running)
any_running = 1; /* some entry is both known and running */
- else if (act) { /* mark it for retry */
+ else if (act) {
+ /* Mark-for-close all TLS connections to this node, since
+ * otherwise there could be one that started 30 seconds
+ * ago, and in 30 seconds it will time out, causing us to mark
+ * the node down and undermine the retry attempt. We mark even
+ * the established conns, since if the network just came back
+ * we'll want to attach circuits to fresh conns. */
+ connection_or_set_bad_connections(ri->cache_info.identity_digest, 1);
+
+ /* mark it for retry */
router_set_status(ri->cache_info.identity_digest, 1);
e->can_retry = 1;
e->bad_since = 0;
diff --git a/src/or/circuituse.c b/src/or/circuituse.c
index ee1705b..f651ef7 100644
--- a/src/or/circuituse.c
+++ b/src/or/circuituse.c
@@ -955,8 +955,15 @@ circuit_build_failed(origin_circuit_t *circ)
* to blame, blame it. Also, avoid this relay for a while, and
* fail any one-hop directory fetches destined for it. */
const char *n_conn_id = circ->cpath->extend_info->identity_digest;
+ int already_marked = 0;
if (circ->_base.n_conn) {
or_connection_t *n_conn = circ->_base.n_conn;
+ if (n_conn->is_bad_for_new_circs) {
+ /* no need to blow away circuits/streams/etc. Also, don't mark this
+ * router as newly down, since maybe this was just an old circuit
+ * attempt that's finally timing out now. */
+ already_marked = 1;
+ }
log_info(LD_OR,
"Our circuit failed to get a response from the first hop "
"(%s:%d). I'm going to try to rotate to a better connection.",
@@ -966,7 +973,7 @@ circuit_build_failed(origin_circuit_t *circ)
log_info(LD_OR,
"Our circuit died before the first hop with no connection");
}
- if (n_conn_id) {
+ if (n_conn_id && !already_marked) {
entry_guard_register_connect_status(n_conn_id, 0, 1, time(NULL));
/* if there are any one-hop streams waiting on this circuit, fail
* them now so they can retry elsewhere. */
diff --git a/src/or/connection_or.c b/src/or/connection_or.c
index 6b648b1..836e7fa 100644
--- a/src/or/connection_or.c
+++ b/src/or/connection_or.c
@@ -610,7 +610,7 @@ connection_or_get_for_extend(const char *digest,
* appropriate. Helper for connection_or_set_bad_connections().
*/
static void
-connection_or_group_set_badness(or_connection_t *head)
+connection_or_group_set_badness(or_connection_t *head, int force)
{
or_connection_t *or_conn = NULL, *best = NULL;
int n_old = 0, n_inprogress = 0, n_canonical = 0, n_other = 0;
@@ -622,8 +622,9 @@ connection_or_group_set_badness(or_connection_t *head)
if (or_conn->_base.marked_for_close ||
or_conn->is_bad_for_new_circs)
continue;
- if (or_conn->_base.timestamp_created + TIME_BEFORE_OR_CONN_IS_TOO_OLD
- < now) {
+ if (force ||
+ or_conn->_base.timestamp_created + TIME_BEFORE_OR_CONN_IS_TOO_OLD
+ < now) {
log_info(LD_OR,
"Marking OR conn to %s:%d as too old for new circuits "
"(fd %d, %d secs old).",
@@ -718,8 +719,10 @@ connection_or_group_set_badness(or_connection_t *head)
}
}
-/** Go through all the OR connections, and set the is_bad_for_new_circs
+/** Go through all the OR connections (or if <b>digest</b> is non-NULL, just
+ * the OR connections with that digest), and set the is_bad_for_new_circs
* flag on:
+ * - all connections if <b>force</b> is true.
* - all connections that are too old.
* - all open non-canonical connections for which a canonical connection
* exists to the same router.
@@ -732,13 +735,14 @@ connection_or_group_set_badness(or_connection_t *head)
* better than another.
*/
void
-connection_or_set_bad_connections(void)
+connection_or_set_bad_connections(const char *digest, int force)
{
if (!orconn_identity_map)
return;
DIGESTMAP_FOREACH(orconn_identity_map, identity, or_connection_t *, conn) {
- connection_or_group_set_badness(conn);
+ if (!digest || !memcmp(digest, conn->identity_digest, DIGEST_LEN))
+ connection_or_group_set_badness(conn, force);
} DIGESTMAP_FOREACH_END;
}
diff --git a/src/or/connection_or.h b/src/or/connection_or.h
index 7176302..216a9bd 100644
--- a/src/or/connection_or.h
+++ b/src/or/connection_or.h
@@ -18,7 +18,7 @@ or_connection_t *connection_or_get_for_extend(const char *digest,
const tor_addr_t *target_addr,
const char **msg_out,
int *launch_out);
-void connection_or_set_bad_connections(void);
+void connection_or_set_bad_connections(const char *digest, int force);
int connection_or_reached_eof(or_connection_t *conn);
int connection_or_process_inbuf(or_connection_t *conn);
diff --git a/src/or/main.c b/src/or/main.c
index b59351c..582a1c2 100644
--- a/src/or/main.c
+++ b/src/or/main.c
@@ -1173,7 +1173,7 @@ run_scheduled_events(time_t now)
circuit_expire_old_circuits_serverside(now);
/** 5. We do housekeeping for each connection... */
- connection_or_set_bad_connections();
+ connection_or_set_bad_connections(NULL, 0);
for (i=0;i<smartlist_len(connection_array);i++) {
run_connection_housekeeping(i, now);
}
--
1.7.1