Hello!

Retain messages are not sending after several days uptime (random count of days).
Restarting of plugin make it works again. And I could not see specific errors in logs.

Only receiving last retained message when subscriber application restarted.

OS: Windows
EMQ Broker version : 2.3.10
Eclipse paho

When I am publishing 10 messages on same topic, during that time for some reason subscriber application crashed, after sometime sub application restarted, I am getting only last retaining message only and losing all other data which is big problem.

I am putting Retained flag is true , clean session flag as false and Qos as 1 on both pub and sub.

can some one help on this.
Am I missing any configuration or Erlang broker behaves like getting only last retaining message.

Where are the messages stored when selecting disc option in storage_type? Does it get stored on ram and later moved to disc or does it overflow to disc beyond some limit?

I have loaded this module in EMQTT broker, all retained messages get lost when I restart the EMQTT broker.

Following is the configuration file for emq-mod-retainer module I am using:

module.retainer.storage_type = disc


module.retainer.max_message_num = 100000


module.retainer.max_payload_size = 64KB


module.retainer.expired_after = 0

We try to use emqtt in kubernetes with glusterFs by using storage_type=disc but folder is created differently as 'emqtt@' in restart of service because Ip address changes.
So we try to give IP address as 0.0.0.0 so that it would be same after every restart but still we cannot use retain folder after reboot,
Is there something we are doing wrong or is there a way to reuse retain messages on restart of emqtt

We're now using the Global Level configuration retainer.expiry_interval to configure the TTL of a retained message.

But does EMQ support something like Message Level configuration about that? I want to configure TTL per message.

I'm still getting retained messages lost with 2.3.4 EMQ on broker restart.
My configuration:

##--------------------------------------------------------------------
## Retainer Module
##--------------------------------------------------------------------

## Where to store the retained messages.
## Notice that all nodes in a cluster are to have the same storage_type.
##
## Value: ram | disc | disc_only
##  - ram: memory only
##  - disc: both memory and disc
##  - disc_only: disc only
##
## Default: ram
retainer.storage_type = disc

## Maximum number of retained messages allowed.
##
## Value: Number >= 0
retainer.max_message_num = 1000000

## Maximum payload size of a retained message.
##
## Value: Bytes
retainer.max_payload_size = 64KB

## Expiration interval of the retained messages. Never expire if the value is 0.
##
## Value: Duration
##  - h: hour
##  - m: minute
##  - s: second
##
## Examples:
##  - 2h:  2 hours
##  - 30m: 30 minutes
##  - 20s: 20 seconds
##
## Defaut: 0
retainer.expiry_interval = 0

Where are the retained messages stored on a disk? Can we copy/backup these files and use it on another broker instance ?

I'd like to fetch all retained message from a topic pattern. Can you guys consider implement it either through management API or cli ?

I have retained mqtt messages on disk. I would like to delete all of these messages. Please suggest a way to do this. Thanks

I wish I had found it a year ago.

Can you add a little more to the documentation?

How does this work? How does it share data across RAM and hard drive.
If the server crashes, is the data backed up?
Is there any api? Can I edit the data files?
Are there data files.

Right now it just appears to me to be a black box, so I have no judgement on how it performs,
how well it distributes, what its limitations are.

Tough to decide if I should use it or not.

I am tempted to use topic tracing to a RAM disk. How does this ocmpare?