packaging
/
ostree
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fetcher: Drop max queue size assertion in libsoup/libcurl backends 

Since f4d1334e19 the primary pull code maintains a
maximum queue. In that commit message I said `Note that I kept an assertion.`.
But I think this is wrong since while it covers a lot of the normal cases, if
one is e.g. trying to fetch a ton of refs, the primary pull code doesn't yet
queue those.  While it'd be nice to queue those, it isn't worth carrying
extra assertions in the backends that can still trigger.

Closes: https://github.com/ostreedev/ostree/issues/1451

Closes: #1453
Approved by: dbnicholson
This commit is contained in:
Colin Walters 2018-02-14 15:28:17 -05:00 committed by Atomic Bot
parent 8dd68fb9f9
commit 96eec98f39
2 changed files with 8 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/libostree/ostree-fetcher-curl.c
View File

@ -866,12 +866,6 @@ _ostree_fetcher_request_async (OstreeFetcher *self,
initiate_next_curl_request (req, task);
g_hash_table_add (self->outstanding_requests, g_steal_pointer (&task));
/* Sanity check, I added * 2 just so we don't abort if something odd happens,
* but we do want to abort if we're asked to do obviously too many requests.
*/
g_assert_cmpint (g_hash_table_size (self->outstanding_requests), <,
_OSTREE_MAX_OUTSTANDING_FETCHER_REQUESTS * 2);
}
void

14
src/libostree/ostree-fetcher-soup.c
View File

@ -60,7 +60,6 @@ typedef struct {
int base_tmpdir_dfd;
GVariant *extra_headers;
int max_outstanding;
gboolean transfer_gzip;
/* Our active HTTP requests */
@ -386,8 +385,6 @@ start_pending_request (ThreadClosure *thread_closure,
OstreeFetcherPendingURI *pending;
GCancellable *cancellable;
g_assert_cmpint (g_hash_table_size (thread_closure->outstanding), <, thread_closure->max_outstanding);
pending = g_task_get_task_data (task);
cancellable = g_task_get_cancellable (task);
@ -473,7 +470,6 @@ ostree_fetcher_session_thread (gpointer data)
{
ThreadClosure *closure = data;
g_autoptr(GMainContext) mainctx = g_main_context_ref (closure->main_context);
gint max_conns;
/* This becomes the GMainContext that SoupSession schedules async
* callbacks and emits signals from. Make it the thread-default
@ -494,17 +490,23 @@ ostree_fetcher_session_thread (gpointer data)
/* XXX: Now that we have mirrorlist support, we could make this even smarter
* by spreading requests across mirrors. */
gint max_conns;
g_object_get (closure->session, "max-conns-per-host", &max_conns, NULL);
if (max_conns < _OSTREE_MAX_OUTSTANDING_FETCHER_REQUESTS)
{
/* We download a lot of small objects in ostree, so this
* helps a lot. Also matches what most modern browsers do. */
* helps a lot. Also matches what most modern browsers do.
*
* Note since https://github.com/ostreedev/ostree/commit/f4d1334e19ce3ab2f8872b1e28da52044f559401
* we don't do queuing in this libsoup backend, but we still
* want to override libsoup's currently conservative
* #define SOUP_SESSION_MAX_CONNS_PER_HOST_DEFAULT 2 (as of 2018-02-14).
*/
max_conns = _OSTREE_MAX_OUTSTANDING_FETCHER_REQUESTS;
g_object_set (closure->session,
"max-conns-per-host",
max_conns, NULL);
}
closure->max_outstanding = 3 * max_conns;
/* This model ensures we don't hit a race using g_main_loop_quit();
* see also what pull_termination_condition() in ostree-repo-pull.c