From d184e6123f400aa1e65f251133935465733562ff Mon Sep 17 00:00:00 2001 From: Lisandro Pin Date: Thu, 12 Feb 2026 00:40:36 +0100 Subject: [PATCH] Document volume server maintenance mode for https://github.com/seaweedfs/seaweedfs/issues/7977 --- Volume-Management.md | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/Volume-Management.md b/Volume-Management.md index 5fb0f2d..65a8b36 100644 --- a/Volume-Management.md +++ b/Volume-Management.md @@ -66,3 +66,26 @@ The balancing plan will try to evenly spread the number of writable and readonly Run `weed shell` and `volume.mount -node : -volumeId ` to mount a volume file. To mount all new volume files you can send a hang-up signal to the volume server causing a reload with a command such as `pkill -HUP -f "weed volume"`. + +## Servicing live volumes + +When dealing with hardware storage issues, it can be useful to prevent writes to volume servers _without_ stopping the service altogether - f.ex. on volumes with RAID storage backends. Volume servers support a maintenance mode for this: when enabled, the server becomes read-only. Reads will succeed, but any write attempt will error out. + +Maintenance mode can be managed via the `volumeServer.state` shell command: + +``` +> volumeServer.state +192.168.10.111:9007 -> Maintenance mode: no +192.168.10.111:9008 -> Maintenance mode: no +192.168.10.111:9009 -> Maintenance mode: no + +> volumeServer.state --nodes 192.168.10.111:9009 --maintenanceOn +192.168.10.111:9009 -> Maintenance mode: yes + +> volumeServer.state +192.168.10.111:9007 -> Maintenance mode: no +192.168.10.111:9008 -> Maintenance mode: no +192.168.10.111:9009 -> Maintenance mode: yes +``` + +Maintenance mode is a sticky server state flag. Changes are effective immediately, and will persist even if the server is restarted. \ No newline at end of file