Delivery messages
- Last UpdatedApr 24, 2025
- 1 minute read
The PubSub Engine uses messages for all delivery tasks in the system. These messages are reference-counted, and their correct management is crucial to the correct operation of the engine.
The Local Delivery Layer (LDL) sends out debug information about the messages in the engine at one-second intervals. The following is published for topics related to delivery messages:
pubsub.debug.messages.<debugItem>
Where <debugItem> can be any of the options in the table below:
|
Debug Item |
Description |
|---|---|
|
free_list_count |
Number of messages allocated but not used. |
|
free_list_<#>_count |
Newer versions of the PubSub Engine will publish additional details on the distribution of the free list. The unused messages are placed on one of four free lists, based of the size of the message payload. List1 contains messages with the smallest payloads, growing progressively until List4, which contains the largest payloads. A normal distribution would have the most messages on List1, and progressively fewer messages on each subsequent list. The sum of the number of messages on these four lists should equal the total published on the pubsub.debug.messages.free_list_count topic. |
|
in_use_count |
Number of messages actively being used. |
|
allocated |
Number of messages allocated. |
|
ndl_queue |
Number of messages cached for re-transmits over the network. |
|
app_queues |
Number of messages awaiting delivery to local applications. |
|
topic_cache.percent_in_use |
The percent used of the topic cache. |
|
topic_cache.references |
Number of references in the topic cache. |
|
topic_cache.size |
Number of items in the topic cache. |
|
security_group_cache.percent_in_use |
The percent used of the security group cache. |
|
security_group_cache.references |
Number of references in the security group cache. |
|
security_group_cache.size |
Number of items in the security group cache. |