[sheepdog-users] how to maintain the consistency of write partial fail and how to update object cache of gateway ?

[Hitoshi Mitake]

At Tue, 4 Aug 2015 11:50:31 +0800,
sky_cord at 163.com wrote:
> 
> [1  <multipart/alternative (7bit)>]
> [1.1  <text/plain; ISO-8859-1 (base64)>]
> 
> 
> 
> 
> 
> 
> Hi,    I read the sheepdog source code recently, and have some questions as below :    1) write process is strong consistency, but if one node write ok, two fail during write flow, how to maintain the         consistency of the three objects ?    2) how to update the object cache of the gateway  if objects transfered to other nodes when one or more nodes down.        during this condition, object content already modified, but object cache does not update
>      thanks

Hi, I'd like to answer.

1. the case means down of VM or node. You can recover consistency with "dog vdi check <vdi name>".
2. object cache is not maintained currently. Please do not use it.

Thanks,
Hitoshi

> 
> 
> 
> sky_cord at 163.com
> 
> [1.2  <text/html; ISO-8859-1 (quoted-printable)>]
> 
> [2  <text/plain; us-ascii (7bit)>]
> -- 
> sheepdog-users mailing lists
> sheepdog-users at lists.wpkg.org
> https://lists.wpkg.org/mailman/listinfo/sheepdog-users