Also, frankly, the multi doesn't otherwise have any way of communicating
the contents obtained by get to the subsequent create.
This alternative sequence is much cleaner that way.
On Fri, Jan 11, 2013 at 7:30 PM, Marshall McMullen <
[EMAIL PROTECTED]> wrote:
> Actually I thought about this some more and one workaround you could do I
> suppose is do the Get outside of the Multi and use that for the Create/Set
> in the Multi followed by a Delete against the old path with a version
> check. That way if the old path was modified and its versioned changed the
> multi will fail and not delete the old data.
> So, I think this is entirely feasible without waiting on 1407. Let me know
> if you want some more pointers on how to do it.
> On Fri, Jan 11, 2013 at 5:41 PM, Himanshu Vashishtha
> <[EMAIL PROTECTED]>wrote:
> > Thanks for the info Marshall; will watch 1407.
> > On Fri, Jan 11, 2013 at 4:23 PM, Marshall McMullen <
> > [EMAIL PROTECTED]> wrote:
> > > Multi currently doesn't support the ability to encode a Get operation
> > into
> > > the multi. So there's no way to atomically move data from one znode to
> > > another. There is a patch in progress to add this support to Multi,
> > > would make this feasible. See:
> > > https://issues.apache.org/jira/browse/ZOOKEEPER-1407
> > >
> > > On Fri, Jan 11, 2013 at 5:19 PM, Himanshu Vashishtha
> > > <[EMAIL PROTECTED]>wrote:
> > >
> > > > Pardon in case this question has been asked before: I am wondering
> > about
> > > > the possibility/feasibility of move/rename a znode in zookeeper.
> > > > Now that multi() is in, it looks like it should be doable, but I may
> > > > wrong. Is there any plans/jiras to support this.
> > > >
> > > > Thanks,
> > > > Himanshu
> > > >
> > >