Reloading after creating a collection

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Reloading after creating a collection

gnandre
Hi,

Why is it that I need to reload collection after I created it? CDCR runs
into issues if I do not do this.
Reply | Threaded
Open this post in threaded view
|

Re: Reloading after creating a collection

Shawn Heisey-2
On 9/19/2019 9:36 AM, Arnold Bronley wrote:
> Why is it that I need to reload collection after I created it? CDCR runs
> into issues if I do not do this.

If the config doesn't change after creation, I would not expect that to
be required.

If you do change the config to enable CDCR after the collection is
created, then you have to reload so that Solr sees the new config.

Thanks,
Shawn
Reply | Threaded
Open this post in threaded view
|

Re: Reloading after creating a collection

gnandre
Hi,
I am not changing the the config to enable CDCR. I am just using the CDCR
API to start it. Does that count as changing configuration?

On Thu, Sep 19, 2019 at 12:20 PM Shawn Heisey <[hidden email]> wrote:

> On 9/19/2019 9:36 AM, Arnold Bronley wrote:
> > Why is it that I need to reload collection after I created it? CDCR runs
> > into issues if I do not do this.
>
> If the config doesn't change after creation, I would not expect that to
> be required.
>
> If you do change the config to enable CDCR after the collection is
> created, then you have to reload so that Solr sees the new config.
>
> Thanks,
> Shawn
>
Reply | Threaded
Open this post in threaded view
|

Re: Reloading after creating a collection

Shawn Heisey-2
On 9/19/2019 12:09 PM, Arnold Bronley wrote:
> I am not changing the the config to enable CDCR. I am just using the CDCR
> API to start it. Does that count as changing configuration?

I would guess that there are no changes to the config from using the
API, but I also admit that I have never tried CDCR.  It would strike me
as very unusual to have that change the config, but I didn't write CDCR.

That might be something we should treat as a bug.  Before opening an
issue in Jira, would you be able to come up with precise steps and a
minimal config to reproduce?

Thanks,
Shawn