Details
-
Bug
-
Resolution: Fixed
-
Critical
-
2.6.0
-
None
-
None
-
1
Description
This is a bug introduced in 2.6.0 which will always add a timeout operator to ops with durability requirements. This is always ok (> 0) but only if the async API is used without an explicit timeout we call .timeout() with 0 so the timer will fire very quickly. This needs to be fixed since it lets API calls like this time out prematurely: bucket.async().upsert(doc, PersistTo.MASTER); .. note the async(), on sync it's fine. Workaround for now is to always apply an explicit timeout as in bucket.async().upsert(doc, PersistTo.MASTER, 5, TimeUnit.SECONDS);