mahara-contributors team mailing list archive

[Aaron Wells <1427885@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

Originally, we just had each artefact store its parent ID. This is slow
because it requires running multiple queries to find all the descendants
of a node.

Then, we added a "path" element to each artefact. This is better, but
you can't get a performance improvement by indexing the column, because
most of the queries rely on the "LIKE" operator. (See
https://bugs.launchpad.net/mahara/+bug/1423700 )

So if we want to squeeze more performance out of this, I think the one
remaining thing to look into is the "nested set" technique. This
technique results in very fast searches for descendants, with the cost
of somewhat slower writes. http://mikehillyer.com/articles/managing-
hierarchical-data-in-mysql/

There are even existing PHP libraries for using the technique, such as
this one: http://www.sideralis.org/baobab/

** Affects: mahara
     Importance: Wishlist
         Status: Triaged


** Tags: performance

** Changed in: mahara
   Importance: Undecided => Wishlist

** Changed in: mahara
    Milestone: None => 15.10.0

-- 
You received this bug notification because you are a member of Mahara
Contributors, which is subscribed to Mahara.
Matching subscriptions: Subscription for all Mahara Contributors -- please ask on #mahara-dev or mahara.org forum before editing or unsubscribing it!
https://bugs.launchpad.net/bugs/1427885

Title:
  Change "artefact.path" column to use the "nested set" technique for
  managing hierarchical data

Status in Mahara ePortfolio:
  Triaged

Bug description:
  Originally, we just had each artefact store its parent ID. This is
  slow because it requires running multiple queries to find all the
  descendants of a node.

  Then, we added a "path" element to each artefact. This is better, but
  you can't get a performance improvement by indexing the column,
  because most of the queries rely on the "LIKE" operator. (See
  https://bugs.launchpad.net/mahara/+bug/1423700 )

  So if we want to squeeze more performance out of this, I think the one
  remaining thing to look into is the "nested set" technique. This
  technique results in very fast searches for descendants, with the cost
  of somewhat slower writes. http://mikehillyer.com/articles/managing-
  hierarchical-data-in-mysql/

  There are even existing PHP libraries for using the technique, such as
  this one: http://www.sideralis.org/baobab/

To manage notifications about this bug go to:
https://bugs.launchpad.net/mahara/+bug/1427885/+subscriptions