nchan: Out of shared memory while allocating message of size 26053. Increase nchan_max_reserved_memory.

August 29, 2025Aug 29

Server locking with the following - diagnostics attached:

Aug 28 15:20:41 Angband nginx: 2025/08/28 15:20:41 [crit] 18252#18252: ngx_slab_alloc() failed: no memory

Aug 28 15:20:41 Angband nginx: 2025/08/28 15:20:41 [error] 18252#18252: shpool alloc failed

Aug 28 15:20:41 Angband nginx: 2025/08/28 15:20:41 [error] 18252#18252: nchan: Out of shared memory while allocating message of size 26053. Increase nchan_max_reserved_memory.

Aug 28 15:20:41 Angband nginx: 2025/08/28 15:20:41 [error] 18252#18252: *173999 nchan: error publishing message (HTTP status code 500), client: unix:, server: , request: "POST /pub/devices?buffer_length=1 HTTP/1.1", host: "localhost"

Aug 28 15:20:41 Angband nginx: 2025/08/28 15:20:41 [error] 18252#18252: MEMSTORE:01: can't create shared message for channel /devices

Aug 28 15:20:41 Angband webgui: Nchan out of shared memory. Reloading nginx

Aug 28 15:20:41 Angband rc.nginx: Restarting Nginx server daemon...

Aug 28 15:20:41 Angband rc.nginx: Checking configuration for correct syntax and then trying to open files referenced in configuration...

Aug 28 15:20:41 Angband rc.nginx: /usr/sbin/nginx -t -c /etc/nginx/nginx.conf

Aug 28 15:20:41 Angband rc.nginx: Stopping Nginx server daemon gracefully...

Aug 28 15:20:53 Angband nginx: 2025/08/28 15:20:53 [error] 18252#18252: MEMSTORE:01: force-reaping msg with refcount 1

angband-diagnostics-20250828-2210.zip

Quote

August 29, 2025Aug 29

Community Expert

@Squid you wanted to know some things if this was encountered again.

My understanding that this was patched / fixed.

I mad a script found at the bottom of this forum regarding similar
https://forums.unraid.net/topic/182004-nchan_max_reserved_memory-errors-varlog-full-what-to-do/#findComment-1496660

please file a bug report. here
https://forums.unraid.net/bug-reports/stable-releases/

As this is kinda a known issues but still not 100% reproducible. I have yet to see or hit the error personaly.

however, this is a new error I haven't seen from nginx:
nginx: 2025/08/28 15:20:53 [error] 18252#18252: MEMSTORE:01: force-reaping msg with refcount 1

*The Nginx error message "MEMSTORE:01: force-reaping msg with refcount 1" indicates an issue related to Nginx's internal memory management, specifically within a component or module that utilizes a memory store.

Explanation of the error:

MEMSTORE:01:
This likely refers to a specific memory store or a particular instance of a memory management component within Nginx. The ":01" might indicate a specific type or identifier for that store.
force-reaping msg:
This suggests that Nginx is actively attempting to reclaim memory associated with a "message" or data structure. "Force-reaping" implies that this action is being taken due to a necessity, possibly to prevent memory exhaustion or to clean up resources that are no longer properly referenced.
with refcount 1:
This is a crucial part of the error. "Refcount" (reference count) indicates how many pointers or references currently point to a particular piece of data in memory. A refcount of 1 means that only one entity is currently referencing this data. When Nginx is "force-reaping" a message with a refcount of 1, it implies that the data should ideally have been deallocated or handled by the single entity referencing it, but for some reason, Nginx is stepping in to forcibly reclaim the memory.

Possible Causes:

This error often points to:

Memory Leak or Improper Resource Deallocation:
A module or process within Nginx might be failing to properly release memory or deallocate resources when they are no longer needed.
Module-Specific Issues:
If you are using third-party Nginx modules, the issue could be within one of those modules, particularly if it's handling data or messages in a way that leads to orphaned memory or incorrect reference counting.
Race Conditions or Concurrency Issues:
In highly concurrent environments, race conditions could lead to situations where a reference count is not correctly decremented, or memory is not released at the appropriate time.
Configuration Errors:
While less common for this specific error, misconfigurations related to memory limits or specific module settings could indirectly contribute to memory management issues.

My First thought is out of Physical Ram. so Page file / other data is needed.

?runnign any user scirpt or editing nginx on the host?

Quote

August 29, 2025Aug 29

Author

5 minutes ago, bmartino1 said:
@Squid you wanted to know some things if this was encountered again.

My understanding that this was patched / fixed.

I mad a script found at the bottom of this forum regarding similar
https://forums.unraid.net/topic/182004-nchan_max_reserved_memory-errors-varlog-full-what-to-do/#findComment-1496660
please file a bug report. here
https://forums.unraid.net/bug-reports/stable-releases/

As this is kinda a known issues but still not 100% reproducible. I have yet to see or hit the error personaly.
however, this is a new error I haven't seen from nginx:
nginx: 2025/08/28 15:20:53 [error] 18252#18252: MEMSTORE:01: force-reaping msg with refcount 1
*The Nginx error message "MEMSTORE:01: force-reaping msg with refcount 1" indicates an issue related to Nginx's internal memory management, specifically within a component or module that utilizes a memory store.
Explanation of the error:
MEMSTORE:01:
This likely refers to a specific memory store or a particular instance of a memory management component within Nginx. The ":01" might indicate a specific type or identifier for that store.
force-reaping msg:
This suggests that Nginx is actively attempting to reclaim memory associated with a "message" or data structure. "Force-reaping" implies that this action is being taken due to a necessity, possibly to prevent memory exhaustion or to clean up resources that are no longer properly referenced.
with refcount 1:
This is a crucial part of the error. "Refcount" (reference count) indicates how many pointers or references currently point to a particular piece of data in memory. A refcount of 1 means that only one entity is currently referencing this data. When Nginx is "force-reaping" a message with a refcount of 1, it implies that the data should ideally have been deallocated or handled by the single entity referencing it, but for some reason, Nginx is stepping in to forcibly reclaim the memory.
Possible Causes:
This error often points to:
Memory Leak or Improper Resource Deallocation:
A module or process within Nginx might be failing to properly release memory or deallocate resources when they are no longer needed.
Module-Specific Issues:
If you are using third-party Nginx modules, the issue could be within one of those modules, particularly if it's handling data or messages in a way that leads to orphaned memory or incorrect reference counting.
Race Conditions or Concurrency Issues:
In highly concurrent environments, race conditions could lead to situations where a reference count is not correctly decremented, or memory is not released at the appropriate time.
Configuration Errors:
While less common for this specific error, misconfigurations related to memory limits or specific module settings could indirectly contribute to memory management issues.
My First thought is out of Physical Ram. so Page file / other data is needed.

?runnign any user scirpt or editing nginx on the host?

I'm not running any scripts or editing the nginx directly.

Bug report filed, I'll take a look at your script, thanks!

Quote

1

August 30, 2025Aug 30

Thanks @bmartino1 Already replied in the bug report, and probably that is the best place to continue it

Quote

1

nchan: Out of shared memory while allocating message of size 26053. Increase nchan_max_reserved_memory.

Featured Replies

Join the conversation

Account

Navigation

Search

Configure browser push notifications

Chrome (Android)

Chrome (Desktop)

Safari (iOS 16.4+)

Safari (macOS)

Edge (Android)

Edge (Desktop)

Firefox (Android)

Firefox (Desktop)