Re: Potential regression: high server-side memory consumption during import

Branko Čibej <brane_at_apache.org> writes:

> So if I understand this debate correctly: The authz code is so much
> faster now that parsing the authz file and performing the authz lookups
> beats calculating its MD5 checksum?

More that reading/checksumming is still too slow to be done repeatedly.

1.9 reads the file once, per connection, and then does authz lookups on
lots of paths. The authz rules are fixed for the duration of the
connection.

1.10 was reading and performing the checksum repeatedly as well as doing
the authz lookups on lots of paths. The authz rules can change during
the connection lifetime. The authz lookups are faster than 1.9 but not
enough to offset the repeated reading/checksumming.

1.11 goes back to reading the file once, and still does the same authz
lookups. The authz rules once again remain fixed for the duration of
the connection.

-- 
Philip