[sheepdog-users] 答复: the sheeddog is closed unexceptedly, anyone can help me ?

[redtone]

The corosync run properly.
I can run sheep to re-enable sheepdog.

-----邮件原件-----
发件人: Liu Yuan [mailto:namei.unix at gmail.com] 
发送时间: 2013年11月18日 12:48
收件人: redtone
抄送: sheepdog-users at lists.wpkg.org
主题: Re: [sheepdog-users] the sheeddog is closed unexceptedly, anyone can
help me ?

On Mon, Nov 18, 2013 at 12:35:31PM +0800, redtone wrote:
> The following is the log 
> 
>  
> 
> Nov 17 10:09:43   WARN [gway 18469] wait_forward_request(409) poll timeout
> 1, disks of some nodes or network is busy. Going to poll-wait again
> 
> Nov 17 10:09:43   WARN [gway 18149] wait_forward_request(409) poll timeout
> 1, disks of some nodes or network is busy. Going to poll-wait again
> 
> Nov 17 10:09:45   WARN [gway 18435] wait_forward_request(409) poll timeout
> 1, disks of some nodes or network is busy. Going to poll-wait again
> 
> Nov 17 19:22:14  EMERG [main] cdrv_cpg_confchg(573) PANIC: Network
partition
> is detected
> 

Either your switch or corosync is borken. I'd suggest you to switch to
zookeeper
if possible, which would be more stable and reliable than corosync.

Thanks
Yuan