4.0 Strange Commit/Replication Issue

[Briggs Thompson <w....@gmail.com>]

Hello all,

I am running 4.0 alpha and have encountered something I am unable to
explain. I am indexing content to a master server, and the data is
replicating to a slave. The odd part is that when searching through the UI,
no documents show up on master with a standard *:* query. All cache types
are set to zero. I know indexing is working because I am watching the logs
and I can see documents getting added, not to mention the data is written
to the filesystem. I have autocommit set to 60000 (1 minute) so it isn't a
commit issue.

The very strange part is that the slave is correctly replicating the data,
and it is searchable in the UI on the slave (but not master). I don't
understand how/why the data is visible on the slave and not visible on the
master. Does anyone have any thoughts on this or seen it before?

Thanks in advance!
Briggs

Re: 4.0 Strange Commit/Replication Issue

[Briggs Thompson <w....@gmail.com>]

That is the problem. I wasn't aware of that new feature in 4.0. Thanks for
the quick response Tomás.

-Briggs

On Wed, Aug 1, 2012 at 3:08 PM, Tomás Fernández Löbbe <tomasflobbe@gmail.com
> wrote:

> Could your autocommit in the master be using "openSearcher=false"? If you
> go to the Master admin, do you see that the searcher has all the segments
> that you see in the filesystem?
>
>
>
> On Wed, Aug 1, 2012 at 4:24 PM, Briggs Thompson <
> w.briggs.thompson@gmail.com
> > wrote:
>
> > Hello all,
> >
> > I am running 4.0 alpha and have encountered something I am unable to
> > explain. I am indexing content to a master server, and the data is
> > replicating to a slave. The odd part is that when searching through the
> UI,
> > no documents show up on master with a standard *:* query. All cache types
> > are set to zero. I know indexing is working because I am watching the
> logs
> > and I can see documents getting added, not to mention the data is written
> > to the filesystem. I have autocommit set to 60000 (1 minute) so it isn't
> a
> > commit issue.
> >
> > The very strange part is that the slave is correctly replicating the
> data,
> > and it is searchable in the UI on the slave (but not master). I don't
> > understand how/why the data is visible on the slave and not visible on
> the
> > master. Does anyone have any thoughts on this or seen it before?
> >
> > Thanks in advance!
> > Briggs
> >
>

Re: 4.0 Strange Commit/Replication Issue

[Tomás Fernández Löbbe <to...@gmail.com>]

Could your autocommit in the master be using "openSearcher=false"? If you
go to the Master admin, do you see that the searcher has all the segments
that you see in the filesystem?



On Wed, Aug 1, 2012 at 4:24 PM, Briggs Thompson <w.briggs.thompson@gmail.com
> wrote:

> Hello all,
>
> I am running 4.0 alpha and have encountered something I am unable to
> explain. I am indexing content to a master server, and the data is
> replicating to a slave. The odd part is that when searching through the UI,
> no documents show up on master with a standard *:* query. All cache types
> are set to zero. I know indexing is working because I am watching the logs
> and I can see documents getting added, not to mention the data is written
> to the filesystem. I have autocommit set to 60000 (1 minute) so it isn't a
> commit issue.
>
> The very strange part is that the slave is correctly replicating the data,
> and it is searchable in the UI on the slave (but not master). I don't
> understand how/why the data is visible on the slave and not visible on the
> master. Does anyone have any thoughts on this or seen it before?
>
> Thanks in advance!
> Briggs
>