mahara-contributors team mailing list archive

[Nigel Cunningham <nigelc@xxxxxxxxxxxxxxx>]

After having done some profiling and preparation of patches over the
last few days, I think 32M is reasonable for normal operation, but we
will definitely still need to raise the memory limit for some
operations.

The main one I've focused on so far is adding users from a CSV file.
Profiling got me focusing on execution time as well as memory use, and I
have a few patches that together do better than halving the execution
time (one in particular - to handle_event - will probably have far wider
implications and accounts for most of the savings) while only raising
memory use by a small amount. 64MB is more than enough even for an
import of 5000 users.

-- 
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/785472

Title:
  Eliminate all raise_memory_limit() calls

Status in Mahara ePortfolio:
  Triaged

Bug description:
  There are very few places in Mahara were we raise PHP's memory limit,
  but it is bad practice since that memory may not be reclaimed by
  Apache until that child process is terminated.

  Unless we are confident that a particular operation is impossible to
  do without raising the memory limit, we should rewrite code which
  fails to work within 32 MB of memory.

To manage notifications about this bug go to:
https://bugs.launchpad.net/mahara/+bug/785472/+subscriptions