replica count documentation

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Fri, Apr 15, 2011 at 07:09:26AM -0700, Gally, Jerry wrote:
> Does that imply that we can create/modify file elements on one system,
> halt that system, and not see the new/modified file elements on a second
> system unless the second system had scanned its mount point before the
> first system went down?

As I understand it it depends on how you've got the one system mounted. If
you're mounting through Gluster, the theory of it is that any
creation/modification goes to both systems at once, so the situation you're
worried about could happen, but only if the file creation/modification was
in the middle of happening and had not completed on one system at the time
the system it hadn't completed on went down. That would be when triggering
the "healing" is required to sync them later.

So the way you've put the question, about halting the first system, no that
shouldn't block the replication - unless you'd also mounted the first system
_outside_ of Gluster, and wrote/modified the file on it directly. That's
where the confusion came in that started this thread. 

At least that's what I make of it.

Whit


[Index of Archives]     [Gluster Development]     [Linux Filesytems Development]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux