Excessive File Watchers Cause "too many open files" Error in openimserver-openim-push and openimserver-openim-rpc-friend Pods

[cubxxw]

Body:

Description

When monitoring the logs of the openimserver-openim-push and openimserver-openim-rpc-friend pods, the following error was encountered after the service started:

failed to create fsnotify watcher: too many open files

The error seems to be related to reaching the maximum file watchers limit.

Logs

For openimserver-openim-push pod:

configFolderPath: use config ../config/config.yaml use config ../config/notification.yaml PortFromConfig: port PortFromConfig: prometheus_port start openimserver-openim-push server, port: 80 prometheusPort: 0 , OpenIM version: v3.2.0 failed to create fsnotify watcher: too many open files

And for openimserver-openim-rpc-friend pod:

configFolderPath: use config ../config/config.yaml use config ../config/notification.yaml PortFromConfig: port PortFromConfig: prometheus_port start openimserver-openim-rpc-friend server, port: 80 prometheusPort: 0 , OpenIM version: v3.2.0

Workaround

The issue was temporarily mitigated by increasing the system file watcher limits with the following commands:

sudo sysctl -w fs.inotify.max_user_watches=2099999999
sudo sysctl -w fs.inotify.max_user_instances=2099999999
sudo sysctl -w fs.inotify.max_queued_events=2099999999

However, this might not be a sustainable solution, especially in environments with multiple running pods/services. It is critical to understand why so many files are being watched and whether optimizations can be applied to reduce the number.

Environment

• OpenIM version: v3.2.0
• Kubernetes version: [please specify]
• Cloud provider or hardware configuration: [please specify]

Expected Behavior

The services/pods should run without encountering the "too many open files" error and without requiring modifications to system-level file watcher settings.

Potential Solutions

• Investigate the necessity of the number of active file watchers and explore code-level optimizations.
• Consider providing guidelines or configurations for handling scenarios where high numbers of file watchers are required.

[kubbot]

This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 7 days.

[cubxxw]

restart

[kubbot]

This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 7 days.

[kubbot]

This issue was closed because it has been stalled for 7 days with no activity.