Split caching out from the stasis_caching_topic.

In working with res_stasis, I discovered a significant limitation to
the current structure of stasis_caching_topics: you cannot subscribe
to cache updates for a single channel/bridge/endpoint/etc.

To address this, this patch splits the cache away from the
stasis_caching_topic, making it a first class object. The stasis_cache
object is shared amongst individual stasis_caching_topics that are
created per channel/endpoint/etc. These are still forwarded to global
whatever_all_cached topics, so their use from most of the code does
not change.

In making these changes, I noticed that we frequently used a similar
pattern for bridges, endpoints and channels:

     single_topic  ---------------->  all_topic
           ^
           |
     single_topic_cached  ----+---->  all_topic_cached
                              |
                              +---->  cache

This pattern was extracted as the 'Stasis Caching Pattern', defined in
stasis_caching_pattern.h. This avoids a lot of duplicate code between
the different domain objects.

Since the cache is now disassociated from its upstream caching topics,
this also necessitated a change to how the 'guaranteed' flag worked
for retrieving from a cache. The code for handling the caching
guarantee was extracted into a 'stasis_topic_wait' function, which
works for any stasis_topic.

(closes issue ASTERISK-22002)
Review: https://reviewboard.asterisk.org/r/2672/


git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@395954 65c4cc65-6c06-0410-ace0-fbb531ad65f3
This commit is contained in:
David M. Lee
2013-08-01 13:49:34 +00:00
parent 5c13969469
commit e1b959ccbb
54 changed files with 1179 additions and 448 deletions

View File

@@ -73,6 +73,28 @@ ASTERISK_FILE_VERSION(__FILE__, "$Revision$")
</managerEvent>
***/
static struct stasis_cp_all *endpoint_cache_all;
struct stasis_cp_all *ast_endpoint_cache_all(void)
{
return endpoint_cache_all;
}
struct stasis_cache *ast_endpoint_cache(void)
{
return stasis_cp_all_cache(endpoint_cache_all);
}
struct stasis_topic *ast_endpoint_topic_all(void)
{
return stasis_cp_all_topic(endpoint_cache_all);
}
struct stasis_topic *ast_endpoint_topic_all_cached(void)
{
return stasis_cp_all_topic_cached(endpoint_cache_all);
}
static struct ast_manager_event_blob *peerstatus_to_ami(struct stasis_message *msg);
STASIS_MESSAGE_TYPE_DEFN(ast_endpoint_snapshot_type);
@@ -80,10 +102,6 @@ STASIS_MESSAGE_TYPE_DEFN(ast_endpoint_state_type,
.to_ami = peerstatus_to_ami,
);
static struct stasis_topic *endpoint_topic_all;
static struct stasis_caching_topic *endpoint_topic_all_cached;
static struct ast_manager_event_blob *peerstatus_to_ami(struct stasis_message *msg)
{
struct ast_endpoint_blob *obj = stasis_message_data(msg);
@@ -168,16 +186,6 @@ void ast_endpoint_blob_publish(struct ast_endpoint *endpoint, struct stasis_mess
}
}
struct stasis_topic *ast_endpoint_topic_all(void)
{
return endpoint_topic_all;
}
struct stasis_caching_topic *ast_endpoint_topic_all_cached(void)
{
return endpoint_topic_all_cached;
}
struct ast_endpoint_snapshot *ast_endpoint_latest_snapshot(const char *tech,
const char *name, unsigned int guaranteed)
{
@@ -190,8 +198,12 @@ struct ast_endpoint_snapshot *ast_endpoint_latest_snapshot(const char *tech,
return NULL;
}
msg = stasis_cache_get_extended(ast_endpoint_topic_all_cached(),
ast_endpoint_snapshot_type(), id, guaranteed);
if (guaranteed) {
stasis_topic_wait(ast_endpoint_topic_all_cached());
}
msg = stasis_cache_get(ast_endpoint_cache(),
ast_endpoint_snapshot_type(), id);
if (!msg) {
return NULL;
}
@@ -267,44 +279,28 @@ struct ast_json *ast_endpoint_snapshot_to_json(
return ast_json_ref(json);
}
static void endpoints_stasis_shutdown(void)
static void endpoints_stasis_cleanup(void)
{
stasis_caching_unsubscribe_and_join(endpoint_topic_all_cached);
endpoint_topic_all_cached = NULL;
STASIS_MESSAGE_TYPE_CLEANUP(ast_endpoint_snapshot_type);
STASIS_MESSAGE_TYPE_CLEANUP(ast_endpoint_state_type);
ao2_cleanup(endpoint_topic_all);
endpoint_topic_all = NULL;
ao2_cleanup(endpoint_cache_all);
endpoint_cache_all = NULL;
}
int ast_endpoint_stasis_init(void)
{
ast_register_atexit(endpoints_stasis_shutdown);
int res = 0;
ast_register_cleanup(endpoints_stasis_cleanup);
if (STASIS_MESSAGE_TYPE_INIT(ast_endpoint_snapshot_type) != 0) {
endpoint_cache_all = stasis_cp_all_create("endpoint_topic_all",
endpoint_snapshot_get_id);
if (!endpoint_cache_all) {
return -1;
}
if (!endpoint_topic_all) {
endpoint_topic_all = stasis_topic_create("endpoint_topic_all");
}
res |= STASIS_MESSAGE_TYPE_INIT(ast_endpoint_snapshot_type);
res |= STASIS_MESSAGE_TYPE_INIT(ast_endpoint_state_type);
if (!endpoint_topic_all) {
return -1;
}
if (!endpoint_topic_all_cached) {
endpoint_topic_all_cached =
stasis_caching_topic_create(
endpoint_topic_all, endpoint_snapshot_get_id);
}
if (!endpoint_topic_all_cached) {
return -1;
}
if (STASIS_MESSAGE_TYPE_INIT(ast_endpoint_state_type) != 0) {
return -1;
}
return 0;
return res;
}