Bug 701784 - Receiver dequeue rate slower when acking messages in batches greater than receiver capacity
Summary: Receiver dequeue rate slower when acking messages in batches greater than rec...
Keywords:
Status: NEW
Alias: None
Product: Red Hat Enterprise MRG
Classification: Red Hat
Component: qpid-cpp
Version: 1.3
Hardware: Unspecified
OS: Unspecified
medium
unspecified
Target Milestone: ---
: ---
Assignee: messaging-bugs
QA Contact: MRG Quality Engineering
URL:
Whiteboard:
Depends On:
Blocks: 698367 794748
TreeView+ depends on / blocked
 
Reported: 2011-05-03 20:28 UTC by Andy Goldstein
Modified: 2020-11-04 18:29 UTC (History)
1 user (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed:
Target Upstream Version:


Attachments (Terms of Use)

Description Andy Goldstein 2011-05-03 20:28:19 UTC
Description of problem: When a Receiver has a capacity (e.g. 10) that is less than the number of messages acknowledged in a batch (e.g. 11), the Receiver's dequeue rate drops significantly.


Version-Release number of selected component (if applicable): latest MRG-M 1.3 and 2.0 prerelease RPMs


How reproducible: 100%


Steps to Reproduce:
1. Send messages to a queue
2. Receive messages from queue using get(), with Receiver capacity 10
3. Call session.acknowledge() after every 11th message received
  
Actual results: on my server, can only receive somewhere between 250 and 600 messages/sec


Expected results: much more than 500 or so messages/second.  With capacity 10, batch size 10, I can receive several thousand messages/sec.


Note You need to log in before you can comment on or make changes to this bug.