Hi Martin,
Zebra is Koha's indexing engine. Check the manual for documentation
about the daemon or the cron job for rebuilds. In most production
instances, Zebra is updated very frequently. Full rebuilds are needed
rarely.
Koha can now also use Elastic Search though this is still experimental.
Stick with Zebra for now.
Plack is middleware that improves Koha's performance. As Mirko said, it
has nothing to do with Zebra.
So there has been some confusion in the information the vendor gave you,
or that you received. Never mind.
If however that vendor is making changes and not contributing those
changes to Koha, that is not unique in the world, but it should give you
pause. It implies the vendor is making the software available as a
cloud service only. If they are distributing the software, they must
provide the source code or they will be in breach of Koha's license.
From the viewpoint of a library choosing a vendor, the relevant
question is whether you want to be part of a world wide community-driven
project, or with a vendor who takes the output of the community's effort
but thinks they'll get some advantage from keeping their own efforts
secret. That route raises a number of risks for libraries including
using software that has not passed Koha's rather rigorous QA process and
ultimately perhaps becoming locked in to the vendor's fork.
Each library makes that choice for themselves. There is a list of
vendors at koha-community.org though the project does not endorse
these. Make your own inquiries including checking Koha's About page (in
the staff client) Release Notes, meeting minutes and mailing lists to
see who is contributing to Koha and speaking with customers of the vendor.
Regards,
Bob Birchall
Calyx
On 14/12/16 04:25, Martin House wrote:
Well, darn. I asked the question because a third party vendor I contacted
suggested that plack could replace zebra as the overall index. Of course,
I'm investigating moving our self-hosted system over to this company. :)
Shows you how much I know. (or how much this vendor knows). That said, the
vendor did say they have developed koha on their own, and made
modifications. I just assumed that I could do the same, with plack as they
have.
That's a bummer. I may need to see about setting up cron jobs that will
call zebra-rebuild daily. I tried that but have not been very successful.
Thanks (Vielen Dank!)
Martin
Martin House
martin.ho...@gmail.com
www.martindhouse.com
704-996-0297
"An unexamined life is not worth living."
On Tue, Dec 13, 2016 at 12:19 PM, Mirko Tietgen <mi...@abunchofthings.net>
wrote:
Hi,
Martin House schrieb am 13.12.2016
Is there anything I should be aware of, before moving my production
server
from zebra to plack?
That question does not work. Plack does not index your data as Zebra
does. You can't do what you describe here.
In other words, have any of you run in to issues that
may not be noted in the official instructions? My follow up question is,
do
I need to remove zebra after plack install, or do I just leave it sitting
there?
You cannot remove Zebra or Koha will break. You cannot exchange
Plack for Zebra, because Plack does not do what Zebra does.
I've looked at a few sites, and can't find concrete answers to these
questions.
Try asking the right questions. Are you really talking about Plack?
Are you maybe thinking about Elasticsearch instead?
PS. I'm sick of having to run zebra-reindex every day, manually, so I
figure it's time to investigate plack.
Why would you manually run Zebra? Use the indexer daemon or at least
a cron job.
HTH,
Mirko
_______________________________________________
Koha mailing list http://koha-community.org
Koha@lists.katipo.co.nz
https://lists.katipo.co.nz/mailman/listinfo/koha
_______________________________________________
Koha mailing list http://koha-community.org
Koha@lists.katipo.co.nz
https://lists.katipo.co.nz/mailman/listinfo/koha
_______________________________________________
Koha mailing list http://koha-community.org
Koha@lists.katipo.co.nz
https://lists.katipo.co.nz/mailman/listinfo/koha
