Issue checklist
Please put x inside of the box that matches your issue.
Describe the bug
We are experiencing recurring severe database-related hitches and full svMain hangs across multiple FiveM servers using oxmysql.
This is currently affecting three separate servers:
- 2 servers running Ubuntu/Linux
- 1 server running Windows
The issue presents similarly across all three environments.
The servers will run normally for a period of time and then begin producing large server thread hitch warnings. In some cases, oxmysql queries take multiple seconds to complete immediately before or during the hitch.
Example:
[script:oxmysql] qbx_core took 2767.3559ms to execute a query!
followed immediately by:
[citizen-server-impl] server thread hitch warning: timer interval of 2763 milliseconds
We have also seen much larger server-thread hitches ranging from approximately 15 to 60+ seconds.
Eventually the server can stop responding entirely and report:
Error: Loop svMain seems hung! (last checkin 45 seconds ago)
txAdmin then stops receiving heartbeats and restarts the server.
We have also repeatedly seen:
[script:oxmysql] Warning: Failed to retrieve latest version of oxmysql (ETIMEDOUT).
and:
[citizen-server-impl] Server list query returned an error: EOF
The svMain watchdog does not consistently report the same resource. It has pointed at multiple unrelated resources across different hangs, which makes it appear that the named resource is simply where the main thread happened to be when it stalled.
Example watchdog output:
svMain watchdog stack: qbx_binoculars: event monitor:cb_... <- qbx_adminmenu: event monitor:exists <- qbx_binoculars: tick <- root
This behavior has now been observed on multiple independent servers and operating systems.
Screenshots
Screenshots are attached showing:
- oxmysql queries taking multiple seconds
- server thread hitch warnings occurring at the same time
svMain watchdog hangs
- txAdmin heartbeat loss and automatic restarts
Code
Example query shown in the logs:
SELECT userId, citizenid, license, name, charinfo, money, job, gang, position, metadata,
UNIX_TIMESTAMP(last_logged_out) AS lastLoggedOutUnix
FROM players
WHERE citizenid = ?
<img width="1315" height="180" alt="Image" src="https://github.com/user-attachments/assets/2eed4a9c-a3b1-4559-a8ba-447faca5f84e" />
<img width="1315" height="180" alt="Image" src="https://github.com/user-attachments/assets/dea55d23-d5c9-4fd3-9020-35289bfd12b4" />
<img width="993" height="517" alt="Image" src="https://github.com/user-attachments/assets/68a6c2c0-45bf-4bd3-8473-a1a6a0521bd9" />
<img width="1097" height="507" alt="Image" src="https://github.com/user-attachments/assets/bccb6ca7-1474-4015-af8b-8c5aad47c096" />
<img width="977" height="851" alt="Image" src="https://github.com/user-attachments/assets/4b95f037-75b8-43fa-b1f1-7cdf5de6efeb" />
<img width="1456" height="94" alt="Image" src="https://github.com/user-attachments/assets/94235a98-3ab3-4b9a-8125-33219f298327" />
<img width="649" height="409" alt="Image" src="https://github.com/user-attachments/assets/f45d5ab2-a1d6-42f5-b1f7-37be203f379c" />
Issue checklist
Describe the bug
We are experiencing recurring severe database-related hitches and full
svMainhangs across multiple FiveM servers using oxmysql.This is currently affecting three separate servers:
The issue presents similarly across all three environments.
The servers will run normally for a period of time and then begin producing large server thread hitch warnings. In some cases, oxmysql queries take multiple seconds to complete immediately before or during the hitch.
Example:
[script:oxmysql] qbx_core took 2767.3559ms to execute a query!followed immediately by:
[citizen-server-impl] server thread hitch warning: timer interval of 2763 millisecondsWe have also seen much larger server-thread hitches ranging from approximately 15 to 60+ seconds.
Eventually the server can stop responding entirely and report:
Error: Loop svMain seems hung! (last checkin 45 seconds ago)txAdmin then stops receiving heartbeats and restarts the server.
We have also repeatedly seen:
[script:oxmysql] Warning: Failed to retrieve latest version of oxmysql (ETIMEDOUT).and:
[citizen-server-impl] Server list query returned an error: EOFThe
svMainwatchdog does not consistently report the same resource. It has pointed at multiple unrelated resources across different hangs, which makes it appear that the named resource is simply where the main thread happened to be when it stalled.Example watchdog output:
svMain watchdog stack: qbx_binoculars: event monitor:cb_... <- qbx_adminmenu: event monitor:exists <- qbx_binoculars: tick <- rootThis behavior has now been observed on multiple independent servers and operating systems.
Screenshots
Screenshots are attached showing:
svMainwatchdog hangsCode
Example query shown in the logs: