I have a producer which sends persistent messages in /batches/ to a queue
leveraging JMS transaction.
I have tested and found that /Producer Flow Control/ is applied when using a
batch size of 1. I could see my producer being throttled as per the memory
limit I have configured for the queue. Here'
Hi Tim,
Thank you very much for the detailed response.
Yes, I am aware that PFC will kick in eventually and do a collateral damage
while using async sends without configuring /Producer Window Size/.
However, I am not seeing that collateral damage when sending messages in
batches with batch size >
Hi Tim,
The PFC documentation is very clear that PFC won't kick in for async sends
unless /Producer Window Size/ is configured.
/
Producers that use Async Sends - generally speaking, producers of
non-persistent messages - don't bother waiting for any acknowledgement from
the broker; so, if a memo
Could this be a bug? Or am I missing something?
--
View this message in context:
http://activemq.2283324.n4.nabble.com/What-is-the-correct-way-to-throttle-ActiveMQ-producers-who-send-persistent-messages-in-batches-to-a--tp4701204p4701233.html
Sent from the ActiveMQ - User mailing list archive a
Hi Tim,
Thanks once again for taking the time to help me out. I really appreciate
that.
I'm very sorry for not mentioning the ActiveMQ version that I am using.
I'm using ActiveMQ 5.10.0.
Though my concern is about /batch send/ and PFC, I also tested async send to
answer your questions.
I sent aro
Hi Richard,
You told that a consumer consumes messages from a queue and then messages
are published to secondary queues.
Kindly note that PFC will be applied to this publisher.
Are the messages published synchronously or asynchronously to the secondary
queues?
--
View this message in context:
Hi Daniel,
I guess you are getting the exception in your producer and you are sending
messages asynchronously. Is this right?
It would be helpful if you share your activemq.xml configuration here.
--
View this message in context:
http://activemq.2283324.n4.nabble.com/User-Memory-Issues-tp47011
Hi Tim,
I couldn't even send 1000 messages using jms.asyncSend=true (my payload
being 2k) with the storeLimit of 2 mb. And there is no hint whatsoever about
the per-destination memory limit of 1 mb I set. Here are the logs for your
reference.
And here's my activemq configuration (I'm sharing on
I tested again with the storeUsage limit set to 32 mb (Please have a look at
my previous reply to learn about the results I got with 2 mb as storeUsage).
Not much difference. The producer is stopped on hitting the storeUsage
limit.
And here's my relevant activemq configuration:
>From this may
Can someone explain this please? This question was asked in 2011 and yet I
see a similar behaviour with ActiveMQ 5.10.0.
The producer never recovers after the ResourceAllocationException.
There is very little or no documentation related to Producer Window Size.
It would be really helpful to the com
I just realized I was using ActiveMQ 5.8.0 for my tests. I really thought I
was using v5.10.0. And little did I realize that the latest available
version is v5.12.0.
I'll rerun my tests with the latest ActiveMQ version and come back to you
all.
--
View this message in context:
http://activemq.
Sure I must have been living in a cave. Sorry for all the trouble Tim.
As you mentioned, the producer throttling works out of the box while sending
messages in batches. I believe even async sends should be throttled out of
the box without bothering about Producer Window Size.
This should work well
This, I believe has been fixed in 5.9.0. I tested using the latest versions
and found this to be working correctly.
--
View this message in context:
http://activemq.2283324.n4.nabble.com/setProducerWindowSize-inhibits-propagation-of-ResAllocEx-producer-stays-forever-blocked-once-it-blocs-tp3782
Hi Tim,
Yes it's kicking in for the per-destination limit.
It will be great if you can edit the PFC wiki.
But just one question.
Is ProducerWindowSize still relevant?
--
View this message in context:
http://activemq.2283324.n4.nabble.com/What-is-the-correct-way-to-throttle-ActiveMQ-producers
Hi Tim,
Thanks for your continued interest in this subject.
Yes, as per my testing, I didn't configure any ProducerWindowSize but found
that the PFC was kicking in based on the per-destination limit I configured.
I don't think it was the whole-connection PFC because I had set fairly
larger limits
Hi Tim,
Thank you very much for the detailed explanation.
I agree it is not as such bad to see "Producer stopped" message immediately
after "Producer throttled". I merely registered the observations from my
test which had no ProducerWindowSize configured.
Now that my doubts are cleared, let me tr
Here is my finding. With a producer window size of 1 mb (same as my
per-destination memory limit) and a payload of 2 kb both producer and
consumer are eventually stopped after sending fewer than 2k messages while
using a batch size > 1. When I used 100 as the batch size, I get the
following excepti
17 matches
Mail list logo
