[sheepdog-users] md info failed one time

Liu Yuan namei.unix at gmail.com
Tue Jun 18 07:47:33 CEST 2013


On 06/17/2013 05:54 PM, Valerio Pachera wrote:
> Hi, if I get this:
> 
> # collie node md info --all
> Id      Size    Used    Avail   Use%    Path
> Node 0:
>  0      1.8 TB  1.0 TB  777 GB   57%    /mnt/ST2000DM001-1CH164_W1E2N5G6/obj
> Node 1:
>  0      1.8 TB  717 GB  1.1 TB   38%    /mnt/ST2000DM001-1CH164_W1E2N5GM/obj
>  1      465 GB  241 GB  225 GB   51%    /mnt/wd_WMAYP1690412
> Node 2:
> Jun 17 11:37:47 [main] connect_to(254) failed to connect to
> 192.168.6.43:7000: Operation now in progress
> Failed to connect to 192.168.6.43:7000
> 
> is it because the host was busy doing something else?
> I got it only one time.
> 
> I tried then also on the other hosts but it didn't repeat.
> 

This is because of the way we try to solve half-open problem. Basically,
connect(2) time-out for 5 seconds because the server is too busy and
return EINPROGRESS. When you get this error, just retry the command is
sufficient.

Thanks,
Yuan



More information about the sheepdog-users mailing list