3 ms·
I mean… activemq delivers messages once and reliably… we push 12k msgs/s with a gig of ram and 2 cores about 25% cpu. Msg size is about 500 bytes - 500kb. It’d
by exabrial 2y ago
I mean… activemq delivers messages once and reliably… we push 12k msgs/s with a gig of ram and 2 cores about 25% cpu. Msg size is about 500 bytes - 500kb.
It’d be faster but we don’t care. It also participates in an XA transaction…
In 10 years of running it practically non stop up to volumes of 100kmsg/s with hundreds of network and disk issues, I’ve never ever seen even once a dupe message. Sure as the article is it could happen but in practice it doesn’t.
- orf 2y agoYou wouldn’t expect to see issues in a simple system in a steady-state. The problems arise when the system responds to other transient or unexpected issues. Try `kill -9`ing your consumers or unplugging a network cable and see what happens. Also, AMQP is most certainly not exactly once by any measure at all. Do you have any consistency checks or anything to validate duplicates?
- exabrial 2y agoWe don’t use AMQP. And kill -9 is for amateurs. Pull plugs.
- orf 2y agoManually pulling plugs is for amateurs: automate or die Either way, you’re getting duplicate deliveries if you’re using >1 machine. And if you’re not, what are you adding to the conversion?
- exabrial 2y agoUh… we’re not. Sorry to disappoint you.
- orf 2y agoI have never crashed my car, yet I acknowledge it is between a possibility and a certainty, depending on the situation. Saying “Well I’ve driven 100k miles and never crashed” doesn’t add much. Also, not sure how you’re consistency checking 100k messages per second over that many years.