Abount "KahaDB Cleanup" of durable subscribers including pending messages

My understanding is that, KahaDB rewrite the KAHA_SUBSCRIPTION_COMMAND to latest journal file for ease of deletion the oldeset jorunal log, when the durable subscriber has no pending message. But, in the following case, "Durable Subscriber C and D" aren't rewritten to latest journal despite th

the useless commands are written into KahaDB when the durable subscriber is destroyed by "offline timeout"

Hi all, I use ActiveMQ 5.13.1 for MQTT. It seems that ActivMQ writes some useless commands, when the durable subscriber is destroyed by Offline Timeout. For example, the following situation: - Topic a, b, c - Durable Subscriber A, B, C - "Durable Subscriber A" subscribe ONLY "Topic a" -

Re: It costs time when stopping ActiveMQ

Thanks for reply. I haven't investgate JVM yet. I'm using AMQ as follows conditions, is there a factor that is likely to take time? Conditions:  * Using MQTT (QoS=2)  * MQTT Client is Eclipse Paho 1.0.0  * Subscriber's CleanSession is false.  * Number of Subscribers is 1 over.  * Number of To

It costs time when stopping ActiveMQ

I am now using ActiveMQ 5.13.1. When I tried to stop ActiveMQ, it cost more than 20 minutes. By checking the activemq.log, it looks like that "uptime XX minutes" were not outputted and it costs time when trying to stop KahaDB. Normal:   INFO | Stopping async topic tasks   INFO | Stopped KahaDB

Why is the index inconsistent with the journal files in KahaDB?

The old journal files aren't cleanup when running ActiveMQ under the following conditions. Conditions: * Using ActiveMQ 5.13.1 * Using MQTT (QoS=2) * MQTT Client is Eclipse Paho 1.0.0 * Subscriber's CleanSession is false. * Number of Subscribers is 1000 over. * Number of Topics is 1000 over. * A T