diff options
| author | Sascha Roland <sroland@ngncc.de> | 2018-08-29 17:53:47 +0200 |
|---|---|---|
| committer | antirez <antirez@gmail.com> | 2018-09-04 13:13:36 +0200 |
| commit | c1e9186f069e0b3aa1507e0e86a16d5001ea5420 (patch) | |
| tree | 37545214b296f0a41f99a6ac0fd3350c039cefef | |
| parent | d60c17cbb3ae0a4b1a435f0b6fc986a60449ebec (diff) | |
| download | redis-c1e9186f069e0b3aa1507e0e86a16d5001ea5420.tar.gz | |
#5299 Fix blocking XREAD for streams that ran dry
The conclusion, that a xread request can be answered syncronously in
case that the stream's last_id is larger than the passed last-received-id
parameter, assumes, that there must be entries present, which could be
returned immediately.
This assumption fails for empty streams that actually contained some
entries which got removed by xdel, ... .
As result, the client is answered synchronously with an empty result,
instead of blocking for new entries to arrive.
An additional check for a non-empty stream is required.
| -rw-r--r-- | src/t_stream.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/t_stream.c b/src/t_stream.c index 77fbf4645..40e0c84b8 100644 --- a/src/t_stream.c +++ b/src/t_stream.c @@ -1436,7 +1436,7 @@ void xreadCommand(client *c) { * synchronously in case the group top item delivered is smaller * than what the stream has inside. */ streamID *last = &groups[i]->last_id; - if (streamCompareID(&s->last_id, last) > 0) { + if (s->length && (streamCompareID(&s->last_id, last) > 0)) { serve_synchronously = 1; *gt = *last; } @@ -1444,7 +1444,7 @@ void xreadCommand(client *c) { } else { /* For consumers without a group, we serve synchronously if we can * actually provide at least one item from the stream. */ - if (streamCompareID(&s->last_id, gt) > 0) { + if (s->length && (streamCompareID(&s->last_id, gt) > 0)) { serve_synchronously = 1; } } |