In addition to what tilerator requires, most styles will benefit from custom partial indexes which should be specified by the style.

You want to consider the need to issue statements to

- Create indexes with fillfactor 100 for imports that are not updated, e.g.

CREATE INDEX IF NOT EXISTS planet_osm_roads_admin
  ON planet_osm_roads USING GIST (way)
  WITH (fillfactor=100)
  WHERE boundary = 'administrative';

- Create indexes concurrently e.g.

CREATE INDEX CONCURRENTLY planet_osm_roads_admin
  ON planet_osm_roads USING GIST (way)
  WHERE boundary = 'administrative';

- Reindex e.g.

REINDEX planet_osm_roads_admin;

- Reindex concurrently e.g.

ALTER INDEX planet_osm_roads_admin RENAME TO planet_osm_roads_admin_old;
CREATE INDEX CONCURRENTLY planet_osm_roads_admin
  ON planet_osm_roads USING GIST (way)
  WHERE boundary = 'administrative';
DROP INDEX planet_osm_roads_admin_old;

All of these are tasks which will need to be done by developers, when setting up new databases, or periodically as part of normal maintenance.

Forwarding

Pine

---------- Forwarded message ----------
From: "Guillaume Lederrey" <glederrey@wikimedia.org>
Date: Jun 21, 2016 12:03
Subject: [discovery] Index management for Maps
To: "Operations Engineers" <ops@lists.wikimedia.org>, "A public mailing list about Wikimedia Search and Discovery projects" <discovery@lists.wikimedia.org>
Cc:

Hello!

I need some feedback from my fellow Ops on how to manage indexes on Maps.

Context:

Maps imports OpenStreetMap data in a Postgresql database. This import
is done with osm2pgsql, which takes care of creating the schema,
populating tables and creating a few indexes.

Some additional indexes are required to support the specific
functionalities of Tilerator. So far, those indexes have been created
manually and have not been tracked.

The enhancements to the schema created by osm2pgsql are minor (a few
index and functions), so we probably need a lightweight solution.

Proposal:

A few idempotent scripts are versionned in osm-bright.tm2source [1].
Those scripts are executed after review by Ops, at the request of the
project. We don't use a full schema migration process, because at this
point there isn't really a need for it on this project.

Does this look reasonable to you? Feedback welcomed, shoot me down if
you have to...

Thanks for your help!

   Guillaume


[1] https://github.com/kartotherian/osm-bright.tm2source/tree/master/sql



--
Guillaume Lederrey
Operations Engineer, Discovery
Wikimedia Foundation

_______________________________________________
discovery mailing list
discovery@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/discovery

_______________________________________________
Maps-l mailing list
Maps-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/maps-l